@inbook{
 type = {inbook},
 year = {2025},
 pages = {121-136},
 volume = {15803},
 publisher = {Springer Nature Switzerland},
 city = {Cham},
 id = {970fe68a-423d-327f-94a9-c97b3b633682},
 created = {2025-08-02T01:18:59.073Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T15:19:14.072Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Watabe:HCII:2025},
 private_publication = {false},
 abstract = {In recent years, with societal changes, efforts have been made to design public transportation systems, introduce new mobility options, and establish the necessary legal frameworks. As a result, the demand for public transportation and related services is expected to increase in the future, which also suggests a growing demand for transfer search services. Previous studies have shown that, in addition to the aforementioned factors, ``transfer resistance'' also influences travel route selection. Therefore, this study aims to organize and analyze the contributing factors of transfer resistance, and conduct a simulation of travel route selection based on the model created using these results, in an attempt to provide travel route suggestions that take into account personal preferences.},
 bibtype = {inbook},
 author = {Watabe, Sota and Aihara, Kenro},
 editor = {Streitz, Norbert A. and Konomi, Shin'ichi},
 doi = {10.1007/978-3-031-92980-9_8},
 chapter = {A Model of Transfer Resistance for Personalized Trip in Public Transportation},
 title = {Distributed, Ambient and Pervasive Interactions}
}

In recent years, with societal changes, efforts have been made to design public transportation systems, introduce new mobility options, and establish the necessary legal frameworks. As a result, the demand for public transportation and related services is expected to increase in the future, which also suggests a growing demand for transfer search services. Previous studies have shown that, in addition to the aforementioned factors, ``transfer resistance'' also influences travel route selection. Therefore, this study aims to organize and analyze the contributing factors of transfer resistance, and conduct a simulation of travel route selection based on the model created using these results, in an attempt to provide travel route suggestions that take into account personal preferences.

@article{
 title = {関連性・多様性・意外性チャネルフレームワークによるロングテールアーティスト推薦の強化},
 type = {article},
 year = {2025},
 pages = {286-297},
 volume = {J108-D},
 id = {e33fa167-1a91-3147-9a52-c1d0fd3782fb},
 created = {2025-08-02T15:08:49.545Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:07:47.746Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Sato:IEICE:2025},
 private_publication = {false},
 abstract = {推薦分野において，セレンディピティを考慮した推薦が精度のみを追求した推薦よりもユーザの満足度が上昇することから注目されている．音楽推薦においては人気度バイアスが特に顕著で，Spotifyを調査したRolling Stone誌のデータによれば，上位1%のアーティストが再生数の90%を占めており，深刻なロングテール問題を引き起こしている．またユーザの過去の履歴に類似した推薦は，フィルタバブル現象を引き起こし，ユーザの満足度と予測精度の不一致を招く可能性がある．本研究では，セレンディピティの三つの側面，つまり，関連性，多様性，意外性を考慮したアーティスト推薦法を提案する．提案手法は，マルチチャネル型の自己アテンションを備えたアーティストエンコーダより構成されている．関連性チャネルはアーティスト間の類似性を，多様性チャネルはアーティストのジャンルの多様性を，意外性チャネルではアーティストに関する記述を扱うことで，全体としてセレンディピティの向上を図るとともにロングテール問題の解決を試みる．},
 bibtype = {article},
 author = {佐藤, 諒平 and SITKRONGWONG, Padipat and 相原, 健郎 and 高須, 淳宏},
 doi = {10.14923/transinfj.2024det0008},
 journal = {電子情報通信学会論文誌},
 number = {5}
}

推薦分野において，セレンディピティを考慮した推薦が精度のみを追求した推薦よりもユーザの満足度が上昇することから注目されている．音楽推薦においては人気度バイアスが特に顕著で，Spotifyを調査したRolling Stone誌のデータによれば，上位1%のアーティストが再生数の90%を占めており，深刻なロングテール問題を引き起こしている．またユーザの過去の履歴に類似した推薦は，フィルタバブル現象を引き起こし，ユーザの満足度と予測精度の不一致を招く可能性がある．本研究では，セレンディピティの三つの側面，つまり，関連性，多様性，意外性を考慮したアーティスト推薦法を提案する．提案手法は，マルチチャネル型の自己アテンションを備えたアーティストエンコーダより構成されている．関連性チャネルはアーティスト間の類似性を，多様性チャネルはアーティストのジャンルの多様性を，意外性チャネルではアーティストに関する記述を扱うことで，全体としてセレンディピティの向上を図るとともにロングテール問題の解決を試みる．

@article{
 title = {みんぱくBERT：博物館における展示物の解説文を考慮した追加学習済みBERTモデル},
 type = {article},
 year = {2025},
 pages = {14-23},
 volume = {18},
 websites = {https://ipsj.ixsq.nii.ac.jp/records/2000768},
 id = {4232d114-abbd-32b1-b1a2-1a73d63578b1},
 created = {2025-08-02T15:13:12.367Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:08:36.131Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {本論文では，博物館における展示物の特徴理解に特化したBERTモデルを提案する．BERTは大規模なコーパスで事前学習したモデルを特定のタスクでファインチューニングすることで，様々なタスクに適応できる汎用言語モデルである．近年，ファインチューニングの前に，解きたいタスクに関連するコーパスで追加の学習をすることで，タスクにおける性能が向上することが知られている．このような追加の学習をしたモデルは，医療や法律などのドメインにおいて多く提案されているが，博物館に特化したモデルは提案されていない．博物館においては，鑑賞ルートの推薦や鑑賞観点の獲得などを目的とした研究が行われており，近年ではBERTなどの深層学習モデルを用いて本課題に取り組んでいる．そこで，本研究では，国立民族学博物館の展示物に関する文書データを用いてBERTモデルを追加学習することで，博物館における展示物の特徴理解に特化したBERTモデル（みんぱくBERT）を提案する．みんぱくBERTの有効性を検証するために，みんぱくBERTと事前学習済みBERTを展示物に関する2種類の下流タスクにおいて比較した．結果として，いずれの下流タスクにおいてもみんぱくBERTの有効性が示された．

In this paper, we propose a BERT model specialized for understanding cultural properties in museums. BERT is a general-purpose language model that is pre-trained on a large corpus and can be adapted to a wide range of tasks through fine-tuning. Recent studies have shown that additional learning on a domain-specific corpus before fine-tuning can further improve task performance. While such domain-specific models have been proposed in fields like medicine and law, no models designed specifically for museums have been developed. In the museum domain, research has been conducted on topics such as recommending viewing routes and acquiring viewing perspectives, and recently, deep learning models like BERT have been applied to these tasks. In this study, we propose “MinpakuBERT,” a BERT model specialized for understanding cultural properties in museums, by conducting additional learning using document data from the National Museum of Ethnology. To evaluate its effectiveness, we compared the performance of MinpakuBERT with that of pre-trained BERT on two downstream tasks related to museum exhibits. The results demonstrated that MinpakuBERT outperformed the pre-trained BERT in both tasks.},
 bibtype = {article},
 author = {三林, 亮太 and 相原, 健郎 and 神門, 典子 and 莊司, 慶行 and 山本, 岳洋 and 山本, 祐輔 and 大島, 裕明},
 journal = {情報処理学会論文誌データベース（TOD）},
 number = {1}
}

本論文では，博物館における展示物の特徴理解に特化したBERTモデルを提案する．BERTは大規模なコーパスで事前学習したモデルを特定のタスクでファインチューニングすることで，様々なタスクに適応できる汎用言語モデルである．近年，ファインチューニングの前に，解きたいタスクに関連するコーパスで追加の学習をすることで，タスクにおける性能が向上することが知られている．このような追加の学習をしたモデルは，医療や法律などのドメインにおいて多く提案されているが，博物館に特化したモデルは提案されていない．博物館においては，鑑賞ルートの推薦や鑑賞観点の獲得などを目的とした研究が行われており，近年ではBERTなどの深層学習モデルを用いて本課題に取り組んでいる．そこで，本研究では，国立民族学博物館の展示物に関する文書データを用いてBERTモデルを追加学習することで，博物館における展示物の特徴理解に特化したBERTモデル（みんぱくBERT）を提案する．みんぱくBERTの有効性を検証するために，みんぱくBERTと事前学習済みBERTを展示物に関する2種類の下流タスクにおいて比較した．結果として，いずれの下流タスクにおいてもみんぱくBERTの有効性が示された． In this paper, we propose a BERT model specialized for understanding cultural properties in museums. BERT is a general-purpose language model that is pre-trained on a large corpus and can be adapted to a wide range of tasks through fine-tuning. Recent studies have shown that additional learning on a domain-specific corpus before fine-tuning can further improve task performance. While such domain-specific models have been proposed in fields like medicine and law, no models designed specifically for museums have been developed. In the museum domain, research has been conducted on topics such as recommending viewing routes and acquiring viewing perspectives, and recently, deep learning models like BERT have been applied to these tasks. In this study, we propose “MinpakuBERT,” a BERT model specialized for understanding cultural properties in museums, by conducting additional learning using document data from the National Museum of Ethnology. To evaluate its effectiveness, we compared the performance of MinpakuBERT with that of pre-trained BERT on two downstream tasks related to museum exhibits. The results demonstrated that MinpakuBERT outperformed the pre-trained BERT in both tasks.

@inbook{
 type = {inbook},
 year = {2024},
 pages = {324-338},
 volume = {14718},
 publisher = {Springer Nature Switzerland},
 city = {Cham},
 id = {ce2b1f12-e72d-37c5-89a2-16ffb45c35f5},
 created = {2025-08-02T15:16:35.919Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T15:26:46.410Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Tsuruoka:HCII:2024},
 private_publication = {false},
 abstract = {Electric kick scooters are currently used in many countries, mainly in Europe and the United States. However, the route guidance of electric scooters is used as a substitute for that of bicycles, and it is possible that the routes are not easy to ride. In addition, studies on ride comfort have shown that vertical swaying of a vehicle in relation to the ground has a negative effect on ride comfort, and it is possible that the unevenness of the road surface has an effect on ride comfort for electric scooters as well.},
 bibtype = {inbook},
 author = {Tsuruoka, Ami and Kadokura, Hiroyuki and Shimizu, Tetsuo and Aihara, Kenro},
 editor = {Streitz, Norbert A. and Konomi, Shin'ichi},
 doi = {10.1007/978-3-031-59988-0_20},
 chapter = {The Impact of Road and Urban Infrastructure Variances on Electric Kick Scooter Travel},
 title = {Distributed, Ambient and Pervasive Interactions}
}

Electric kick scooters are currently used in many countries, mainly in Europe and the United States. However, the route guidance of electric scooters is used as a substitute for that of bicycles, and it is possible that the routes are not easy to ride. In addition, studies on ride comfort have shown that vertical swaying of a vehicle in relation to the ground has a negative effect on ride comfort, and it is possible that the unevenness of the road surface has an effect on ride comfort for electric scooters as well.

@inbook{
 type = {inbook},
 year = {2024},
 pages = {272-286},
 volume = {14718},
 websites = {https://link.springer.com/10.1007/978-3-031-59988-0_16},
 publisher = {Springer Nature Switzerland},
 city = {Cham},
 id = {f5110e6e-f016-351f-ba1a-72f6604785d9},
 created = {2025-08-02T15:22:20.450Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:33:09.019Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Makita:HCII:2024},
 private_publication = {false},
 abstract = {It is known that there are various factors that influence people’s intention to go out.
In this study, the effects of each of these factors on the going-out behavior of residents in urban and suburban areas by the range of going out were analyzed and compared. This study focuses on Saitama City. Saitama City is a suburb in the Tokyo Metropolitan area and the convenience of going out varies greatly within the region.
Data of 1,000 respondents to a web-based questionnaire survey are prepared on a variety of factors thought to explain going-out behavior, such as basic attributes, impressions of public transportation, cognitive distance and preferences. And then a random forest classifier was trained by using collected data.
As a result, the study revealed two key findings on the relationship between various factors and intention to go out. Firstly, cognitive distance significantly influences the intention to go out, confirming prior studies. Secondly, two main patterns of going-out behavior emerge: general intention to go out and intention for leisure activities, versus intention to go outside the city or district. Increase of age influences the former pattern, while preferences for longer outings directly impact the latter. This detailed breakdown offers practical insights for local governments to enhance residents’ intention to go out.},
 bibtype = {inbook},
 author = {Makita, So and Aihara, Kenro},
 editor = {Streitz, Norbert A. and Konomi, Shin'ichi},
 doi = {10.1007/978-3-031-59988-0_16},
 chapter = {Toward Facilitating Going Out:on the Relationship Between Residents’ Outing Behavior and Cognitive Distance in Suburbs},
 title = {Distributed, Ambient and Pervasive Interactions}
}

It is known that there are various factors that influence people’s intention to go out. In this study, the effects of each of these factors on the going-out behavior of residents in urban and suburban areas by the range of going out were analyzed and compared. This study focuses on Saitama City. Saitama City is a suburb in the Tokyo Metropolitan area and the convenience of going out varies greatly within the region. Data of 1,000 respondents to a web-based questionnaire survey are prepared on a variety of factors thought to explain going-out behavior, such as basic attributes, impressions of public transportation, cognitive distance and preferences. And then a random forest classifier was trained by using collected data. As a result, the study revealed two key findings on the relationship between various factors and intention to go out. Firstly, cognitive distance significantly influences the intention to go out, confirming prior studies. Secondly, two main patterns of going-out behavior emerge: general intention to go out and intention for leisure activities, versus intention to go outside the city or district. Increase of age influences the former pattern, while preferences for longer outings directly impact the latter. This detailed breakdown offers practical insights for local governments to enhance residents’ intention to go out.

@inproceedings{
 title = {Impressing Museum Visitors by Having Them Answer Questions Toward the Realization of an Interactive Viewing Support System},
 type = {inproceedings},
 year = {2024},
 pages = {137-142},
 month = {3},
 publisher = {IEEE},
 day = {18},
 id = {1daf9730-3db9-3575-86bb-d42a25996d94},
 created = {2025-08-02T15:23:35.068Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:33:46.364Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Morita:ICIET:2024},
 private_publication = {false},
 abstract = {In this study, we examine how museum visitor's experience can be enhanced and made memorable long after their visit by engaging them in interactive question-and-answer sessions about the exhibits. Our hypothesis is that a viewing system in which museum exhibits pose questions to visitors and they respond based on their own interpretations will enable even those accustomed to viewing exhibits to have a more memorable experience. In the experiment, eight subjects viewed exhibits at the National Museum of Ethnology (Minpaku). As part of our process, they answered questions related to the exhibits. Two days later, we conducted follow-up interviews. Experimental results indicate that questions prompting participants to imagine scenarios from their daily lives are more likely to leave lasting impression. Therefore, an interactive question-and-answer approach to viewing exhibits appears to be effective in creating a more memorable museum experience.},
 bibtype = {inproceedings},
 author = {Yuna, Morita and Yamamoto, Takehiro and Shoji, Yoshiyuki and Ohshima, Hiroaki and Yamamoto, Yusuke and Kando, Noriko and Aihara, Kenro},
 doi = {10.1109/ICIET60671.2024.10542822},
 booktitle = {2024 12th International Conference on Information and Education Technology (ICIET)}
}

In this study, we examine how museum visitor's experience can be enhanced and made memorable long after their visit by engaging them in interactive question-and-answer sessions about the exhibits. Our hypothesis is that a viewing system in which museum exhibits pose questions to visitors and they respond based on their own interpretations will enable even those accustomed to viewing exhibits to have a more memorable experience. In the experiment, eight subjects viewed exhibits at the National Museum of Ethnology (Minpaku). As part of our process, they answered questions related to the exhibits. Two days later, we conducted follow-up interviews. Experimental results indicate that questions prompting participants to imagine scenarios from their daily lives are more likely to leave lasting impression. Therefore, an interactive question-and-answer approach to viewing exhibits appears to be effective in creating a more memorable museum experience.

@inproceedings{
 title = {音楽推薦におけるセレンディピティ向上のためのアーティストエンコーダ},
 type = {inproceedings},
 year = {2024},
 websites = {https://confit.atlas.jp/guide/event-img/deim2024/T3-B-6-01/public/pdf?type=in},
 month = {3},
 id = {f113f7c3-6467-3093-ba52-7423572ca90a},
 created = {2025-08-02T15:25:56.403Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:42:28.927Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Sato:DEIM:2024},
 private_publication = {false},
 abstract = {推薦分野において，セレンディピティを考慮した推薦が精度のみを追求した推薦よりもユーザの満足度が上昇することから注目されている．音楽推薦においては人気度バイアスが特に顕著で，Spotify を調査した Rolling Stone誌 1のデータによれば，上位 1 ％のアーティストが再生数の 90 ％を占めており，深刻なロングテール問題を引き起こしている．またユーザの過去の履歴に類似した推薦は，フィルターバブル現象を引き起こし，ユーザの満足度と予測精度の不一致を招く可能性がある．本研究では，セレンディピティを関連性，多様性，意外性の 3 つの要素を定義とし，マルチチャンネル型のアーティスト自己アテンションを提案する．関連性チャンネルではアーティスト間の類似性を，多様性チャンネルではプレイリスト内の未出現アーティストを考慮し，意外性チャンネルではアーティストに関する記述を利用している．これらチャンネルを統合しセレンディピティの向上とロングテール問題の解決を試みる．},
 bibtype = {inproceedings},
 author = {佐藤, 諒平 and Sitkrongwong, Padipat and 相原, 健郎 and 高須, 淳宏},
 booktitle = {第16回 データ工学と情報マネジメントに関するフォーラム 論文集}
}

推薦分野において，セレンディピティを考慮した推薦が精度のみを追求した推薦よりもユーザの満足度が上昇することから注目されている．音楽推薦においては人気度バイアスが特に顕著で，Spotify を調査した Rolling Stone誌 1のデータによれば，上位 1 ％のアーティストが再生数の 90 ％を占めており，深刻なロングテール問題を引き起こしている．またユーザの過去の履歴に類似した推薦は，フィルターバブル現象を引き起こし，ユーザの満足度と予測精度の不一致を招く可能性がある．本研究では，セレンディピティを関連性，多様性，意外性の 3 つの要素を定義とし，マルチチャンネル型のアーティスト自己アテンションを提案する．関連性チャンネルではアーティスト間の類似性を，多様性チャンネルではプレイリスト内の未出現アーティストを考慮し，意外性チャンネルではアーティストに関する記述を利用している．これらチャンネルを統合しセレンディピティの向上とロングテール問題の解決を試みる．

@inproceedings{
 title = {A Gamification System for Acquiring Appreciation Perspectives in Museum},
 type = {inproceedings},
 year = {2024},
 pages = {149-153},
 month = {3},
 publisher = {IEEE},
 day = {18},
 id = {f00fc6dd-a81f-3191-a2ca-b548bc7a78f1},
 created = {2025-08-02T15:30:02.210Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:34:09.647Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Nishimoto:ICIET:2024},
 private_publication = {false},
 abstract = {We find it difficult to remember the artifacts we viewed after visiting the museum. In this paper, we propose a guidance system for museums that aims to make the ar-tifacts more memorable to visitors. This is achieved through the implementation of Mandala Bingo, a game inspired by bingo. The system targets the National Museum of Ethnology (a.k.a. Minpaku) and incorporates gamification into the Minpaku Guide. The objective is to enable users to appreciate the artifacts based on their unique perspectives, making the artifacts more memorable. The evaluation results indicated that visitors became more conscious of the connections between the features of the artifacts. However, one participant mentioned that he became too focused on searching for the artifacts to be used in Bingo. An important future task is to conduct long-term experiments to measure the retention of memories.},
 bibtype = {inproceedings},
 author = {Nishimoto, Kaisei and Aihara, Kenro and Kando, Noriko and Shoji, Yoshiyuki and Yamamoto, Yusuke and Yamamoto, Takehiro and Ohshima, Hiroaki},
 doi = {10.1109/ICIET60671.2024.10542807},
 booktitle = {2024 12th International Conference on Information and Education Technology (ICIET)}
}

We find it difficult to remember the artifacts we viewed after visiting the museum. In this paper, we propose a guidance system for museums that aims to make the ar-tifacts more memorable to visitors. This is achieved through the implementation of Mandala Bingo, a game inspired by bingo. The system targets the National Museum of Ethnology (a.k.a. Minpaku) and incorporates gamification into the Minpaku Guide. The objective is to enable users to appreciate the artifacts based on their unique perspectives, making the artifacts more memorable. The evaluation results indicated that visitors became more conscious of the connections between the features of the artifacts. However, one participant mentioned that he became too focused on searching for the artifacts to be used in Bingo. An important future task is to conduct long-term experiments to measure the retention of memories.

@inbook{
 type = {inbook},
 year = {2023},
 pages = {30-45},
 id = {e4d36caf-ea2c-3fbc-af4d-e9e1c9d4afdb},
 created = {2025-08-02T15:31:37.563Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:35:06.995Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Yu:ICADL:2023},
 private_publication = {false},
 abstract = {This paper proposes a method for enhancing museum visits by turning them into a personalized treasure hunt game. This approach uses gamification to augment visitor experiences, explicitly addressing the issue of retaining information during casual museum visits. This is achieved by creating quizzes customized to each visitor’s interests and providing unique perspectives for appreciating exhibits. These quizzes, developed through analyzing museum guide logs and individual visitor preferences, encourage visitors to explore exhibits, thereby intensifying their engagement proactively. The game entails visitors identifying specific exhibits, guided by hints like images and maps, and inputting the exhibit IDs to win in-game rewards. Field studies employing actual data from the National Museum of Ethnology in Japan evaluated this approach’s efficacy. Experimental results suggest that visitors are more likely to remember exhibit details connected to quiz answers, even two weeks after their visit. Participants also enjoyed the impromptu exploration encouraged by the treasure hunt game.},
 bibtype = {inbook},
 author = {Yu, Jinsong and Takidaira, Shio and Sawaura, Tsukasa and Shoji, Yoshiyuki and Yamamoto, Takehiro and Yamamoto, Yusuke and Ohshima, Hiroaki and Aihara, Kenro and Kando, Noriko},
 doi = {10.1007/978-981-99-8088-8_3},
 chapter = {Personalized Treasure Hunt Game for Proactive Museum Appreciation by Analyzing Guide App Operation Log}
}

This paper proposes a method for enhancing museum visits by turning them into a personalized treasure hunt game. This approach uses gamification to augment visitor experiences, explicitly addressing the issue of retaining information during casual museum visits. This is achieved by creating quizzes customized to each visitor’s interests and providing unique perspectives for appreciating exhibits. These quizzes, developed through analyzing museum guide logs and individual visitor preferences, encourage visitors to explore exhibits, thereby intensifying their engagement proactively. The game entails visitors identifying specific exhibits, guided by hints like images and maps, and inputting the exhibit IDs to win in-game rewards. Field studies employing actual data from the National Museum of Ethnology in Japan evaluated this approach’s efficacy. Experimental results suggest that visitors are more likely to remember exhibit details connected to quiz answers, even two weeks after their visit. Participants also enjoyed the impromptu exploration encouraged by the treasure hunt game.

@article{
 title = {Data-Driven Bridge Weigh-in-Motion},
 type = {article},
 year = {2023},
 pages = {17064-17077},
 volume = {23},
 month = {8},
 day = {1},
 id = {9aed0e41-3d91-3ce0-8496-84defa750500},
 created = {2025-08-02T15:32:45.423Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:35:31.988Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:IEEESensors:2023},
 private_publication = {false},
 abstract = {Heavy vehicle detection in the road system is now an urgent issue from the perspectives of law enforcement and road health monitoring. Weigh-in-motion (WIM) is a technology that estimates vehicle weights without stopping the vehicles. Pavement WIM (PWIM) is expensive and has limited installation locations. Bridge WIM (BWIM), which utilizes bridge components as weight scales, is quite inexpensive and easier to install. BWIM requires the dynamic characteristics of the bridge and traffic conditions for accurate weight estimation. In general, such characteristics are measured by several experimental runs using a vehicle with known axle weights. The weighing accuracy may be greatly degraded due to the influence of the other traveling vehicles. In this article, we propose a data-driven BWIM using a neural network. The model parameters are optimized automatically by video analysis and vehicle identification between WIMs. The model can estimate vehicle weights accurately considering various traffic conditions that may degrade the weighing accuracy.},
 bibtype = {article},
 author = {Kawakatsu, Takaya and Aihara, Kenro and Takasu, Atsuhiro and Nagayama, Tomonori and Adachi, Jun},
 doi = {10.1109/JSEN.2023.3283849},
 journal = {IEEE Sensors Journal},
 number = {15}
}

Heavy vehicle detection in the road system is now an urgent issue from the perspectives of law enforcement and road health monitoring. Weigh-in-motion (WIM) is a technology that estimates vehicle weights without stopping the vehicles. Pavement WIM (PWIM) is expensive and has limited installation locations. Bridge WIM (BWIM), which utilizes bridge components as weight scales, is quite inexpensive and easier to install. BWIM requires the dynamic characteristics of the bridge and traffic conditions for accurate weight estimation. In general, such characteristics are measured by several experimental runs using a vehicle with known axle weights. The weighing accuracy may be greatly degraded due to the influence of the other traveling vehicles. In this article, we propose a data-driven BWIM using a neural network. The model parameters are optimized automatically by video analysis and vehicle identification between WIMs. The model can estimate vehicle weights accurately considering various traffic conditions that may degrade the weighing accuracy.

@inbook{
 type = {inbook},
 year = {2023},
 pages = {18-34},
 volume = {14037},
 publisher = {Springer Nature Switzerland},
 city = {Cham},
 id = {58010ab3-9e72-32bb-adb4-f275a440e1c7},
 created = {2025-08-02T15:34:32.866Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T15:35:10.784Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kaburagi:HCII:2023},
 private_publication = {false},
 abstract = {Luggage is a hassle for travelers. In recent years, various services related to luggage have emerged, as people are increasingly interested in how to handle luggage during travel. Luggage-free is expected to be beneficial for both businesses by expanding their business opportunities, and for travelers by freeing them from their luggage. Previous research on luggage has often focused on limited subjects, such as public transportation. Therefore, this research aims to propose a methodology that allows travelers and other visitors to leave their baggage behind and ease their behavioral restrictions and effectively promote their behavior at their destination. To analyze the influence of luggage, this paper dealt with distances from key stations by using individual behavior logs. The results suggested that tourists' activities may be affected by various constraints such as physical limit, location, temporal, and transportation when they carry luggage. Therefore, carrying luggage may not only restrict one's mobility but also restrict one's choice of places to stay and activities to enjoy. To achieve more freedom in tourism, it is important not only to increase fixed services such as coin lockers but also to improve services such as delivery.},
 bibtype = {inbook},
 author = {Kaburagi, Fumika and Aihara, Kenro},
 editor = {Streitz, Norbert A. and Konomi, Shin'ichi},
 doi = {10.1007/978-3-031-34609-5_2},
 chapter = {Toward Supporting Baggage-Free Walk-Arounds at Travel Destinations: Issues of the Hassle of Baggage During Travel},
 title = {Distributed, Ambient and Pervasive Interactions}
}

Luggage is a hassle for travelers. In recent years, various services related to luggage have emerged, as people are increasingly interested in how to handle luggage during travel. Luggage-free is expected to be beneficial for both businesses by expanding their business opportunities, and for travelers by freeing them from their luggage. Previous research on luggage has often focused on limited subjects, such as public transportation. Therefore, this research aims to propose a methodology that allows travelers and other visitors to leave their baggage behind and ease their behavioral restrictions and effectively promote their behavior at their destination. To analyze the influence of luggage, this paper dealt with distances from key stations by using individual behavior logs. The results suggested that tourists' activities may be affected by various constraints such as physical limit, location, temporal, and transportation when they carry luggage. Therefore, carrying luggage may not only restrict one's mobility but also restrict one's choice of places to stay and activities to enjoy. To achieve more freedom in tourism, it is important not only to increase fixed services such as coin lockers but also to improve services such as delivery.

@inproceedings{
 title = {位置情報を用いたビンゴゲームによる共通点を持つ展示物の推薦},
 type = {inproceedings},
 year = {2023},
 websites = {https://proceedings-of-deim.github.io/DEIM2023/5b-2-4.pdf},
 id = {7ffd2a30-6592-353f-a9fc-1f19b4d77e47},
 created = {2025-08-02T15:37:29.089Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:10:56.943Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Nishimoto:DEIM:2023},
 private_publication = {false},
 abstract = {本論文では，ビンゴゲームをモチーフとした「曼荼羅ビンゴ」を取り入れることによって，博物館や美術館を鑑賞する際に，鑑賞した展示物が，より訪問者の記憶に残るようなガイドシステムを提案する．大阪府にある国立民族学博物館（みんぱく）を対象とし，ガイドシステムである「みんぱくガイド」にゲーミフィケーションの手法を取り入れる．ゲーミフィケーションとは，ゲームの要素をゲーム以外の分野に取り入れることによって，ユーザを内発的に動機づけすることである．博物館の訪問者の興味の対象を，位置情報やデータベースを用いて取得することで，より適切な展示物の提案を行い，実際に使用，評価してもらう．},
 bibtype = {inproceedings},
 author = {西本, 海生 and 相原, 健郎 and 神門, 典子 and 莊司, 慶行 and 山本, 祐輔 and 山本, 岳洋 and 大島, 裕明},
 booktitle = {第15回 データ工学と情報マネジメントに関するフォーラム 論文集}
}

本論文では，ビンゴゲームをモチーフとした「曼荼羅ビンゴ」を取り入れることによって，博物館や美術館を鑑賞する際に，鑑賞した展示物が，より訪問者の記憶に残るようなガイドシステムを提案する．大阪府にある国立民族学博物館（みんぱく）を対象とし，ガイドシステムである「みんぱくガイド」にゲーミフィケーションの手法を取り入れる．ゲーミフィケーションとは，ゲームの要素をゲーム以外の分野に取り入れることによって，ユーザを内発的に動機づけすることである．博物館の訪問者の興味の対象を，位置情報やデータベースを用いて取得することで，より適切な展示物の提案を行い，実際に使用，評価してもらう．

@inproceedings{
 title = {対話型鑑賞システム実現に向けた質問に答えさせることによる博物館鑑賞の印象付け},
 type = {inproceedings},
 year = {2023},
 websites = {https://yamamotolab.net/papers/2023/morita.pdf},
 id = {bc9b9115-a3b8-3f24-8f81-cfd7276388a5},
 created = {2025-08-02T15:39:50.672Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:12:40.423Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Morita:DEIM:2023},
 private_publication = {false},
 abstract = {本研究では，博物館で展示物に関する質問を答えさせることで，鑑賞者の鑑賞体験をしばらくたった後でも印象付けることが可能かを検証する．博物館の展示物側から鑑賞者側に向けて質問を投げかけ，それに鑑賞者が自分の解釈で回答していくといった鑑賞システムがあることで，鑑賞に慣れていない人でも印象に残る鑑賞が可能になると考えた．実験では，国立民族学博物館（みんぱく）の展示物を 8 人の被験者に質問回答しながら鑑賞させ，2 日後にインタビューを行った．実験結果から，展示物が生活にある場面を想像させる質問が印象に残りやすく，質問に答えさせる鑑賞は鑑賞物を印象に残すのに有効であることが示唆された．},
 bibtype = {inproceedings},
 author = {森田, 佑菜 and 山本, 岳洋 and 莊司, 慶行 and 大島, 裕明 and 山本, 祐輔 and 神門, 典子 and 相原, 健郎},
 booktitle = {第15回 データ工学と情報マネジメントに関するフォーラム 論文集}
}

本研究では，博物館で展示物に関する質問を答えさせることで，鑑賞者の鑑賞体験をしばらくたった後でも印象付けることが可能かを検証する．博物館の展示物側から鑑賞者側に向けて質問を投げかけ，それに鑑賞者が自分の解釈で回答していくといった鑑賞システムがあることで，鑑賞に慣れていない人でも印象に残る鑑賞が可能になると考えた．実験では，国立民族学博物館（みんぱく）の展示物を 8 人の被験者に質問回答しながら鑑賞させ，2 日後にインタビューを行った．実験結果から，展示物が生活にある場面を想像させる質問が印象に残りやすく，質問に答えさせる鑑賞は鑑賞物を印象に残すのに有効であることが示唆された．

@inbook{
 type = {inbook},
 year = {2022},
 pages = {127-138},
 volume = {13325},
 publisher = {Springer International Publishing},
 city = {Cham},
 id = {b9594f7f-53e3-3936-b205-f5c76b95d133},
 created = {2025-08-02T15:41:45.624Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T15:42:19.278Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:HCII:2022},
 private_publication = {false},
 abstract = {In order to realize a smart city, it is essential to understand the ever-changing flow of people in the city. Regarding transportation by public transportation, such as trains, bus, and taxis, probe information generated from their operational data of such services are one of the clues, but private cars are operated independently by the citizens and out of such probe data. It is not easy to comprehensively grasp their dynamics because there are also social demands for privacy protection.},
 bibtype = {inbook},
 author = {Aihara, Kenro},
 editor = {Streitz, Norbert A. and Konomi, Shin'ichi},
 doi = {10.1007/978-3-031-05463-1_9},
 chapter = {Understanding Intra-regional Flow of Vehicles Using Automatic License Plate Recognition},
 title = {Distributed, Ambient and Pervasive Interactions. Smart Environments, Ecosystems, and Cities}
}

In order to realize a smart city, it is essential to understand the ever-changing flow of people in the city. Regarding transportation by public transportation, such as trains, bus, and taxis, probe information generated from their operational data of such services are one of the clues, but private cars are operated independently by the citizens and out of such probe data. It is not easy to comprehensively grasp their dynamics because there are also social demands for privacy protection.

@article{
 title = {提示型検索モデルに基づくミュージアム電子ガイドとその利用ログを用いた事前学習・事後学習支援},
 type = {article},
 year = {2022},
 pages = {364-377},
 volume = {63},
 id = {eea8cbe9-f254-3da5-b0de-f4badb052da8},
 created = {2025-08-02T15:44:15.647Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:39:06.754Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Shoji:IPSJ:2022},
 private_publication = {false},
 abstract = {本研究では，提示型検索モデル（Ostensive Search Model）に基づくインタフェースによって鑑賞者個人の興味を反映したミュージアム体験を可能にする電子ガイドを提案し，そのログを分析することで実現可能になった事前学習，事後学習支援システムについても提案する．我々は，国立民族学博物館（みんぱく）の展示物のうち3,053点について，展示物の解説やビデオなどの詳細情報を検索し，メモなどのアノテーションを付与できるiPad用アプリケーションである「みんぱくガイド」を作成した．みんぱくガイドは，鑑賞者が新しい展示物に気付いたり，興味を明確化することができるように，一覧性の高い検索結果画面を中心に情報探索を繰り返せるインタフェースを持っている．このような電子ガイドの操作履歴や位置情報などのログを用いることで，個人のミュージアム体験を色濃く反映した事前・事後学習支援を可能にした．事前学習支援システムでは，ミュージアムに行く前にカードを整理しながら鑑賞計画を立てるウェブアプリケーションの利用を通じて，鑑賞者に自分が何を学びに行くかという鑑賞軸を自覚してもらう．また，事後学習を促す仕組みとして，ログから鑑賞者が興味を持った展示物を推定し，後から鑑賞体験を思い出しやすくするためのパーソナライズされたポストカードを自動生成するシステムも作成した．

This paper proposes an electronic guide application that enables visitors to get a museum experience that reflects their individual interests through an interface based on an Ostensive Search Model. In addition, we propose a pre-learning and post-learning support system that can be connected to our electronic guide. The system is based on a search result screen with a high level of browsability. We propose a system that allows users to search for detailed information such as explanatory texts and videos of exhibits they are interested in and add annotations such as handwritten notes. We created the “Minpaku Guide,” an iPad application that allows users to search for detailed information such as explanatory text and videos on exhibits of interest and add annotations such as notes. We also developed pre-learning support systems connected to our guide application. As a pre-learning support system, we created a web application that allows users to manually organize cards before visiting the museum to clarify what the visitors will learn beforehand. As a post-learning support system, we implemented a system that summarizes the visitor's operation log of the guide application into a postcard that helps to recall the viewing experience later.},
 bibtype = {article},
 author = {莊司, 慶行 and 相原, 健郎 and 大島, 裕明 and 神門, 典子 and 白石, 晃一 and 中島, 悠太 and 山本, 岳洋 and 山本, 祐輔},
 doi = {10.20729/00216243},
 journal = {情報処理学会論文誌},
 number = {3}
}

本研究では，提示型検索モデル（Ostensive Search Model）に基づくインタフェースによって鑑賞者個人の興味を反映したミュージアム体験を可能にする電子ガイドを提案し，そのログを分析することで実現可能になった事前学習，事後学習支援システムについても提案する．我々は，国立民族学博物館（みんぱく）の展示物のうち3,053点について，展示物の解説やビデオなどの詳細情報を検索し，メモなどのアノテーションを付与できるiPad用アプリケーションである「みんぱくガイド」を作成した．みんぱくガイドは，鑑賞者が新しい展示物に気付いたり，興味を明確化することができるように，一覧性の高い検索結果画面を中心に情報探索を繰り返せるインタフェースを持っている．このような電子ガイドの操作履歴や位置情報などのログを用いることで，個人のミュージアム体験を色濃く反映した事前・事後学習支援を可能にした．事前学習支援システムでは，ミュージアムに行く前にカードを整理しながら鑑賞計画を立てるウェブアプリケーションの利用を通じて，鑑賞者に自分が何を学びに行くかという鑑賞軸を自覚してもらう．また，事後学習を促す仕組みとして，ログから鑑賞者が興味を持った展示物を推定し，後から鑑賞体験を思い出しやすくするためのパーソナライズされたポストカードを自動生成するシステムも作成した． This paper proposes an electronic guide application that enables visitors to get a museum experience that reflects their individual interests through an interface based on an Ostensive Search Model. In addition, we propose a pre-learning and post-learning support system that can be connected to our electronic guide. The system is based on a search result screen with a high level of browsability. We propose a system that allows users to search for detailed information such as explanatory texts and videos of exhibits they are interested in and add annotations such as handwritten notes. We created the “Minpaku Guide,” an iPad application that allows users to search for detailed information such as explanatory text and videos on exhibits of interest and add annotations such as notes. We also developed pre-learning support systems connected to our guide application. As a pre-learning support system, we created a web application that allows users to manually organize cards before visiting the museum to clarify what the visitors will learn beforehand. As a post-learning support system, we implemented a system that summarizes the visitor's operation log of the guide application into a postcard that helps to recall the viewing experience later.

@inproceedings{
 title = {MinpakuBERT: A Language Model for Understanding Cultural Properties in Museum},
 type = {inproceedings},
 year = {2022},
 pages = {13-18},
 month = {7},
 publisher = {IEEE},
 id = {9cc4841c-cb1a-3675-86c7-a3c75ebb2520},
 created = {2025-08-02T15:45:06.078Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:40:56.153Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Mibayashi:IIAIAAI:2022},
 private_publication = {false},
 abstract = {In this paper, we propose a BERT model for understanding cultural properties in museum. BERT is a language model used in natural language processing. Generally, BERT is pre-trained on a large corpus and then fine-tuned on a specific task. Recently, additional learning for BERT has been reported to improve task performance in fine tuning. Therefore, we propose a BERT model for understanding cultural properties in museum (MinpakuBERT) by additional learning a BERT model using document data from the National Museum of Ethnology. To evaluate the performance of MinpakuBERT, we compared the performance of MinpakuBERT and pre-trained BERT by fine-tuning them on two tasks related to cultural properties. The two tasks for comparison were classification tasks using the OCM (Outline of Cultural Materials) label for function and the OWC (Outline of World Cultures) label for region. As a result, the performance of MinpakuBERT is higher than the pre-trained BERT model in the two tasks. In addition, the learning speed in fine tuning was also faster for MinpakuBERT than for the pre-trained BERT. The MinpakuBERT was published on a model-sharing website and made available as a language resource for anyone to use.},
 bibtype = {inproceedings},
 author = {Mibayashi, Ryota and Ueta, Masaki and Kawahara, Takafumi and Matsumoto, Naoaki and Yoshimura, Takuma and Aihara, Kenro and Kando, Noriko and Shoji, Yoshiyuki and Nakajima, Yuta and Yamamoto, Takehiro and Yamamoto, Yusuke and Ohshima, Hiroaki},
 doi = {10.1109/IIAIAAI55812.2022.00013},
 booktitle = {2022 12th International Congress on Advanced Applied Informatics (IIAI-AAI)}
}

In this paper, we propose a BERT model for understanding cultural properties in museum. BERT is a language model used in natural language processing. Generally, BERT is pre-trained on a large corpus and then fine-tuned on a specific task. Recently, additional learning for BERT has been reported to improve task performance in fine tuning. Therefore, we propose a BERT model for understanding cultural properties in museum (MinpakuBERT) by additional learning a BERT model using document data from the National Museum of Ethnology. To evaluate the performance of MinpakuBERT, we compared the performance of MinpakuBERT and pre-trained BERT by fine-tuning them on two tasks related to cultural properties. The two tasks for comparison were classification tasks using the OCM (Outline of Cultural Materials) label for function and the OWC (Outline of World Cultures) label for region. As a result, the performance of MinpakuBERT is higher than the pre-trained BERT model in the two tasks. In addition, the learning speed in fine tuning was also faster for MinpakuBERT than for the pre-trained BERT. The MinpakuBERT was published on a model-sharing website and made available as a language resource for anyone to use.

@inproceedings{
 title = {Museum Experience into a Souvenir: Generating Memorable Postcards from Guide Device Behavior Log},
 type = {inproceedings},
 year = {2021},
 pages = {120-129},
 month = {9},
 publisher = {IEEE},
 id = {cf6354ad-c36a-3071-bdbc-5e0f8810aa1d},
 created = {2025-08-02T15:45:43.274Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:41:19.624Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Shoji:JCDL:2021},
 private_publication = {false},
 abstract = {This paper proposes a method for automatically generating postcards that reflect each visitor's museum experience by analyzing the log of our original iPad app that supports and guides personalized navigation in the National Museum of Ethnology. Museum experiences have become more personalized with the evolution of guiding devices. Each visitor views the different exhibits in a different order. Souvenirs serve to remind visitors of their museum experience and cement it in their memories; thus, souvenir postcards should be tailored to each visitor's museum experience. Such tailored postcards can effectively remind visitors of their experiences, deepen their impressions when they look back at them, and promote post-learning. In this paper, we proposed a system that automatically generates a postcard for each visitor for each visit by selecting the five most relevant and impressive exhibits based on the search and navigation logs on our museum guide app. We analyzed the search logs of the guide devices based on the psychological effects on impressions and memory to estimate which exhibits had the strongest impact on the visitors. We then conducted an in laboratory controlled user experiment with 16 participants to check what exhibits had made an impression on visitors by using the implemented system. The results showed that the exhibits that were seen frequently and the exhibits that participants added to their favorites were the most memorable.},
 bibtype = {inproceedings},
 author = {Shoji, Yoshiyuki and Aihara, Kenro and Kando, Noriko and Nakashima, Yuta and Ohshima, Hiroaki and Takidaira, Shio and Ueta, Masaki and Yamamoto, Takehiro and Yamamoto, Yusuke},
 doi = {10.1109/JCDL52503.2021.00024},
 booktitle = {2021 ACM/IEEE Joint Conference on Digital Libraries (JCDL)}
}

This paper proposes a method for automatically generating postcards that reflect each visitor's museum experience by analyzing the log of our original iPad app that supports and guides personalized navigation in the National Museum of Ethnology. Museum experiences have become more personalized with the evolution of guiding devices. Each visitor views the different exhibits in a different order. Souvenirs serve to remind visitors of their museum experience and cement it in their memories; thus, souvenir postcards should be tailored to each visitor's museum experience. Such tailored postcards can effectively remind visitors of their experiences, deepen their impressions when they look back at them, and promote post-learning. In this paper, we proposed a system that automatically generates a postcard for each visitor for each visit by selecting the five most relevant and impressive exhibits based on the search and navigation logs on our museum guide app. We analyzed the search logs of the guide devices based on the psychological effects on impressions and memory to estimate which exhibits had the strongest impact on the visitors. We then conducted an in laboratory controlled user experiment with 16 participants to check what exhibits had made an impression on visitors by using the implemented system. The results showed that the exhibits that were seen frequently and the exhibits that participants added to their favorites were the most memorable.

@inproceedings{
 title = {Image Retrieval by Hierarchy-aware Deep Hashing Based on Multi-task Learning},
 type = {inproceedings},
 year = {2021},
 pages = {486-490},
 month = {8},
 publisher = {ACM},
 day = {24},
 city = {New York, NY, USA},
 id = {e85c1ae6-764a-38d7-b3d2-a810b177ea31},
 created = {2025-08-02T15:47:13.961Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:41:40.808Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Wang:ICMR:2021},
 private_publication = {false},
 abstract = {Deep hashing has been widely used to approximate nearest-neighbor search for image retrieval tasks. Most of them are trained with image-label pairs without any inter-label relationship, which may not make full use of the real-world data. This paper presents deep hashing, named HA2SH, that leverages multiple types of labels with hierarchical structures that an ethnological museum assigns to their artifacts. We experimentally prove that HA2SH can learn to generate hashes that give a better retrieval performance. Our code is available at https://github.com/wbw520/minpaku.},
 bibtype = {inproceedings},
 author = {Wang, Bowen and Li, Liangzhi and Nakashima, Yuta and Yamamoto, Takehiro and Ohshima, Hiroaki and Shoji, Yoshiyuki and Aihara, Kenro and Kando, Noriko},
 doi = {10.1145/3460426.3463586},
 booktitle = {Proceedings of the 2021 International Conference on Multimedia Retrieval}
}

Deep hashing has been widely used to approximate nearest-neighbor search for image retrieval tasks. Most of them are trained with image-label pairs without any inter-label relationship, which may not make full use of the real-world data. This paper presents deep hashing, named HA2SH, that leverages multiple types of labels with hierarchical structures that an ethnological museum assigns to their artifacts. We experimentally prove that HA2SH can learn to generate hashes that give a better retrieval performance. Our code is available at https://github.com/wbw520/minpaku.

@inproceedings{
 title = {Fully-Neural Approach to Vehicle Weighing and Strain Prediction on Bridges Using Wireless Accelerometers},
 type = {inproceedings},
 year = {2021},
 pages = {8027-8031},
 month = {6},
 publisher = {IEEE},
 day = {6},
 id = {cea77ce1-339c-3112-b1f5-024f02cd2142},
 created = {2025-08-02T15:50:31.169Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:42:16.039Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:ICASSP:2021},
 private_publication = {false},
 abstract = {Bridge weigh-in-motion (BWIM) is a technique of estimating vehicle loads on bridges and can be used to assess a bridge’s structural fatigue and therefore its life. BWIM can be realized by analyzing the bridge deflection in terms of its response to moving axle loads. To obtain accurate load estimates, current BWIM systems require strain sensors, whose (re-) installation costs have limited their application. In this paper, we propose a new BWIM approach based on a deep neural network using accelerometers, which are easier to install than strain sensors, thus helping the advancement of low-cost BWIM systems. By learning the bridge dynamism, our model estimates axle loads successfully from the noisy acceleration signals sampled on a real bridge in service.},
 bibtype = {inproceedings},
 author = {Kawakatsu, Takaya and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun and Wang, Haoqi and Nagayama, Tomonori},
 doi = {10.1109/ICASSP39728.2021.9414433},
 booktitle = {ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}
}

Bridge weigh-in-motion (BWIM) is a technique of estimating vehicle loads on bridges and can be used to assess a bridge’s structural fatigue and therefore its life. BWIM can be realized by analyzing the bridge deflection in terms of its response to moving axle loads. To obtain accurate load estimates, current BWIM systems require strain sensors, whose (re-) installation costs have limited their application. In this paper, we propose a new BWIM approach based on a deep neural network using accelerometers, which are easier to install than strain sensors, thus helping the advancement of low-cost BWIM systems. By learning the bridge dynamism, our model estimates axle loads successfully from the noisy acceleration signals sampled on a real bridge in service.

@inbook{
 type = {inbook},
 year = {2021},
 pages = {2961-2969},
 month = {4},
 publisher = {CRC Press},
 day = {19},
 id = {13af89cf-46b7-3e55-a16a-bfb295a63e3e},
 created = {2025-08-02T15:51:05.923Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:42:43.267Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:Bridge:2021},
 private_publication = {false},
 abstract = {When a vehicle crosses a road bridge, the bridge bends and deforms in response. The dynamic characteristics are unique to each bridge and have been utilized for nondestructive health monitoring. Recently, we have developed a comprehensive system for remote monitoring, which is a platform of the Internet of Things era for the inexpensive and efficient monitoring of road infrastructures. In addition, we have proposed modeling techniques for existing bridges based on deep convolutional neural networks (CNNs). Our solution requires less heuristics and no optimization by hand; instead, it requires a camera and several sensors installed on the bridge components. By analyzing traffic situations using surveillance video data, we have collected a large amount of training data about passing vehicles, including traveling speeds, loci, and axle positions. The training data have been utilized for training a deep CNN. The CNN has brought rich information about passing vehicles and bridge dynamics, although the CNN accepts strain data collected by only a single strain gauge. We named this solution deep sensing, and its applications include vehicle weighing (bridge weigh-in-motion) and nondestructive bridge health monitoring based on anomalous signal detection. In actual health monitoring, we need to handle various types of heterogeneity due to constraints such as the deployable locations of strain gauges, as well as the types of bridges. In this paper, we present our experiences in dealing with the heterogeneity and discuss the practical system architecture, including interfaces, subsystems, and requirements for a fully automated modeling system for real road bridges. Our proposal contributes to the progress toward low-cost and compact health monitoring systems.},
 bibtype = {inbook},
 author = {Kawakatsu, T. and Aihara, K. and Takasu, A. and Adachi, J.},
 doi = {10.1201/9780429279119-403},
 chapter = {Deep learning approach to modeling bridge dynamics using cameras and sensors},
 title = {Bridge Maintenance, Safety, Management, Life-Cycle Sustainability and Innovations}
}

When a vehicle crosses a road bridge, the bridge bends and deforms in response. The dynamic characteristics are unique to each bridge and have been utilized for nondestructive health monitoring. Recently, we have developed a comprehensive system for remote monitoring, which is a platform of the Internet of Things era for the inexpensive and efficient monitoring of road infrastructures. In addition, we have proposed modeling techniques for existing bridges based on deep convolutional neural networks (CNNs). Our solution requires less heuristics and no optimization by hand; instead, it requires a camera and several sensors installed on the bridge components. By analyzing traffic situations using surveillance video data, we have collected a large amount of training data about passing vehicles, including traveling speeds, loci, and axle positions. The training data have been utilized for training a deep CNN. The CNN has brought rich information about passing vehicles and bridge dynamics, although the CNN accepts strain data collected by only a single strain gauge. We named this solution deep sensing, and its applications include vehicle weighing (bridge weigh-in-motion) and nondestructive bridge health monitoring based on anomalous signal detection. In actual health monitoring, we need to handle various types of heterogeneity due to constraints such as the deployable locations of strain gauges, as well as the types of bridges. In this paper, we present our experiences in dealing with the heterogeneity and discuss the practical system architecture, including interfaces, subsystems, and requirements for a fully automated modeling system for real road bridges. Our proposal contributes to the progress toward low-cost and compact health monitoring systems.

@inbook{
 type = {inbook},
 year = {2021},
 pages = {2970-2977},
 month = {4},
 publisher = {CRC Press},
 day = {19},
 id = {d554c426-8ab4-3ac9-90f3-0a22e1d1062a},
 created = {2025-08-02T15:52:17.196Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:43:13.674Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:Bridge:2021},
 private_publication = {false},
 abstract = {In order to realize efficient long-term structural health monitoring, this paper proposes an integrated data man-agement platform that integrates support from collecting various data sets to analyzing them to detect state changes and anomalies. This platform is called PIERS, a platform for inexpensive and efficient monitoring of infrastructures. One of features of the proposed system is that the system is compatible with both conventional and IoT approaches.In the conventional method, each component, such as sensor, logger and transmitter, is installed separately and then connected by cables. Loggers convert the output of sensors into a digital value and records it, while trans-mitters send them to the server. On the other hand, the latter approach uses IoT devices that integrate sensor measurement with data recording and transfer. The proposed system gives a solution to collect and manage syn-chronized data efficiently with IoT devices and other sensing modules.The purpose of long-term monitoring is mainly twofold. One is that the road manager can always confirm the status of structures and can promptly be notified by an alert when an anomaly occurs. In addition, it is to be able to look back on the state before and after those anomalies. For this purpose, detailed and high-density data col-lection is not necessary. The other is the purpose used by civil engineering researchers and experts (high-level analysts) to analyze the progress of deterioration of structures. For this purpose, it is required that the measure-ment be accurate and dense. In the proposed system, users including road managers and experts can search data and extracted events with a standard search interface using a web browser and visualize them in a dashboard. In addition, experts in civil engineering can directly access the database through the API from their own analysis programs to analyze phenomena interactively.},
 bibtype = {inbook},
 author = {Aihara, K. and Takasu, A. and Kawakatsu, T. and Kinoshita, A. and Adachi, J.},
 doi = {10.1201/9780429279119-404},
 chapter = {A data management platform for efficient monitoring of infrastructures},
 title = {Bridge Maintenance, Safety, Management, Life-Cycle Sustainability and Innovations}
}

In order to realize efficient long-term structural health monitoring, this paper proposes an integrated data man-agement platform that integrates support from collecting various data sets to analyzing them to detect state changes and anomalies. This platform is called PIERS, a platform for inexpensive and efficient monitoring of infrastructures. One of features of the proposed system is that the system is compatible with both conventional and IoT approaches.In the conventional method, each component, such as sensor, logger and transmitter, is installed separately and then connected by cables. Loggers convert the output of sensors into a digital value and records it, while trans-mitters send them to the server. On the other hand, the latter approach uses IoT devices that integrate sensor measurement with data recording and transfer. The proposed system gives a solution to collect and manage syn-chronized data efficiently with IoT devices and other sensing modules.The purpose of long-term monitoring is mainly twofold. One is that the road manager can always confirm the status of structures and can promptly be notified by an alert when an anomaly occurs. In addition, it is to be able to look back on the state before and after those anomalies. For this purpose, detailed and high-density data col-lection is not necessary. The other is the purpose used by civil engineering researchers and experts (high-level analysts) to analyze the progress of deterioration of structures. For this purpose, it is required that the measure-ment be accurate and dense. In the proposed system, users including road managers and experts can search data and extracted events with a standard search interface using a web browser and visualize them in a dashboard. In addition, experts in civil engineering can directly access the database through the API from their own analysis programs to analyze phenomena interactively.

@inproceedings{
 title = {Location-based Reminder for Memorizing What Visitors Learn at a Museum},
 type = {inproceedings},
 year = {2021},
 pages = {79-87},
 websites = {https://ceur-ws.org/Vol-2863/paper-09.pdf},
 month = {3},
 publisher = {CEUR-WS.org},
 day = {19},
 id = {f4999ecf-ef36-311a-aeb0-c2a9dd2c680d},
 created = {2025-08-02T15:58:06.900Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:14:37.249Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Shoji:CEUR:2021},
 private_publication = {false},
 abstract = {This paper proposes a method to promote a strongly memorable museum experience by sending reminder notifications when visitors approach places related to the exhibits. Memory retention studies have reported that memory that is linked to personal experience is not easily forgotten. Our method estimates real-world places that are related to a museum exhibit by linking characteristic terms that appear in reviews for these places with exhibit commentaries. Relevance between the characteristic terms is calculated by their co-occurrence in documents and web search query suggestions. This enables us to send reminders about exhibits to users’ mobile devices when they approach related locations in their daily life. Through experiments conducted at the National Museum of Ethnology in Japan, participants who received notifications about the exhibits remembered more exhibits one week later.},
 bibtype = {inproceedings},
 author = {Shoji, Yoshiyuki and Aihara, Kenro and D\"urst, Martin J. and Kando, Noriko and Nakaya, Takuya and Ohshima, Hiroaki and Yamamoto, Takehiro and Yamamoto, Yusuke},
 booktitle = {Joint Proceedings of the Second Workshop on Bridging the Gap between Information Science, Information Retrieval and Data Science, and Third Workshop on Evaluation of Personalisation in Information Retrieval}
}

This paper proposes a method to promote a strongly memorable museum experience by sending reminder notifications when visitors approach places related to the exhibits. Memory retention studies have reported that memory that is linked to personal experience is not easily forgotten. Our method estimates real-world places that are related to a museum exhibit by linking characteristic terms that appear in reviews for these places with exhibit commentaries. Relevance between the characteristic terms is calculated by their co-occurrence in documents and web search query suggestions. This enables us to send reminders about exhibits to users’ mobile devices when they approach related locations in their daily life. Through experiments conducted at the National Museum of Ethnology in Japan, participants who received notifications about the exhibits remembered more exhibits one week later.

@inproceedings{
 title = {Development of One-Stop Smart City Application by Interdisciplinary Data Linkage},
 type = {inproceedings},
 year = {2020},
 keywords = {Behavior log,Crowdsourcing,Data linkage,Integration of heterogeneous contents,Location-based service,Smart city},
 pages = {379-390},
 volume = {12203 LNCS},
 publisher = {Springer},
 id = {7688b16a-2355-31b5-b94b-c07b06a883b0},
 created = {2021-10-12T05:23:55.927Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T17:07:34.088Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {The development of open data by local governments and data platforms for each field is progressing. These are broad ranged data on each area, such as traffic, disaster prevention, restaurants and services, and are expected to be useful information sources for citizens and tourists. On the other hand, these data are usually deployed in a network reachable place, but when they have to be handled individually according to its own format, and in some cases, conversion both in format and in semantics are required, which is a barrier to use. By the way, in existing information services for tourists, especially smartphones application services for tourists, the content provided are selective and limited in some specific fields and target areas covered. That is, there is a problem on coverage in content. In addition, there are many cases where has a problem with the cost of maintaining and updating content, and the content may often be obsolete. Given the existence of data platforms developed for each field, the paper presents the data linkage challenges that enable them to be integrated and used. And also the paper shows a one-stop smart city application that has been developed using that function. This application provides information for tourists during normal times, and it can handle the situation in town, such as congestion, in real time in order to grasp the flow and local stagnation that occur in bursts at the time of events or accidents.},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Takasu, Atsuhiro},
 doi = {10.1007/978-3-030-50344-4_27},
 booktitle = {Distributed, Ambient and Pervasive Interactions, LNCS 12203}
}

The development of open data by local governments and data platforms for each field is progressing. These are broad ranged data on each area, such as traffic, disaster prevention, restaurants and services, and are expected to be useful information sources for citizens and tourists. On the other hand, these data are usually deployed in a network reachable place, but when they have to be handled individually according to its own format, and in some cases, conversion both in format and in semantics are required, which is a barrier to use. By the way, in existing information services for tourists, especially smartphones application services for tourists, the content provided are selective and limited in some specific fields and target areas covered. That is, there is a problem on coverage in content. In addition, there are many cases where has a problem with the cost of maintaining and updating content, and the content may often be obsolete. Given the existence of data platforms developed for each field, the paper presents the data linkage challenges that enable them to be integrated and used. And also the paper shows a one-stop smart city application that has been developed using that function. This application provides information for tourists during normal times, and it can handle the situation in town, such as congestion, in real time in order to grasp the flow and local stagnation that occur in bursts at the time of events or accidents.

@inproceedings{
 title = {提示型検索に基づくミュージアム電子ガイドを中核とした事前・事後学習支援},
 type = {inproceedings},
 year = {2020},
 pages = {81-88},
 month = {12},
 publisher = {情報処理学会},
 id = {e4cda9a5-5fc3-3c73-b4d0-85039846b595},
 created = {2025-08-02T16:01:05.360Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:16:39.598Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Shoji:Jinmoncom:2020},
 private_publication = {false},
 abstract = {本研究では，個人が博物館や美術館といったミュージアムを訪れる際に，事前学習を行ったり，事後学習を行ったりすることを支援するシステムの提案を行う．ミュージアムを明確な目的なく漠然と鑑賞することは，本来ミュージアムにおいて可能な体験を十分に満喫することができているといえない．ミュージアムに滞在している時はもちろんのこと，ミュージアムに行く前の事前学習や，ミュージアムに行った後の事後学習が重要になると考えられる．我々は，これまで，国立民族学博物館を対象として，電子ガイドシステムの提案を行ってきた．本研究では，その電子ガイドシステムを中核として，それを利用した事前学習や事後学習の仕組みを提案する．

This paper proposes a system to support both pre- and post-learning when an individual visits a museum. If visitors visit a museum without a clear goal in their mind, they will miss out on a truly meaningful experience that they could have earned from the museum. It is important to learn before visiting a museum and afterwards, as well as during the visit. We have proposed an electronic guide system for the National Museum of Ethnology. In this study, we propose a system for pre- and post-learning using the electronic guide system.},
 bibtype = {inproceedings},
 author = {莊司, 慶行 and 大島, 裕明 and 神門, 典子 and 相原, 健郎 and 白石, 晃一 and 瀧平, 士夫 and 中島, 悠太 and 山本, 岳洋 and 山本, 祐輔 and 楊, 澤華},
 booktitle = {じんもんこん2020論文集}
}

本研究では，個人が博物館や美術館といったミュージアムを訪れる際に，事前学習を行ったり，事後学習を行ったりすることを支援するシステムの提案を行う．ミュージアムを明確な目的なく漠然と鑑賞することは，本来ミュージアムにおいて可能な体験を十分に満喫することができているといえない．ミュージアムに滞在している時はもちろんのこと，ミュージアムに行く前の事前学習や，ミュージアムに行った後の事後学習が重要になると考えられる．我々は，これまで，国立民族学博物館を対象として，電子ガイドシステムの提案を行ってきた．本研究では，その電子ガイドシステムを中核として，それを利用した事前学習や事後学習の仕組みを提案する． This paper proposes a system to support both pre- and post-learning when an individual visits a museum. If visitors visit a museum without a clear goal in their mind, they will miss out on a truly meaningful experience that they could have earned from the museum. It is important to learn before visiting a museum and afterwards, as well as during the visit. We have proposed an electronic guide system for the National Museum of Ethnology. In this study, we propose a system for pre- and post-learning using the electronic guide system.

@inbook{
 type = {inbook},
 year = {2020},
 pages = {1-42},
 publisher = {Springer International Publishing},
 city = {Cham},
 id = {b8bf88a4-3f45-34e6-9bc5-244366f88009},
 created = {2025-08-02T16:02:06.199Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:44:29.602Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:HandbookSmartCities:2020},
 private_publication = {false},
 abstract = {The chapter describes initiatives aimed at improving the quality and efficiency of urban life by understanding the urban situation by utilizing an unspecified number of citizens and sharing collected and aggregated information, mainly using commoditized devices such as smartphones. Specifically, the chapter focuses on efforts to improve the efficiency of road management work by collecting road traffic conditions using a drive recorder application and acquiring sensor data that can estimate the road surface condition. It also shows a crowdsourced bus location service that can be realized at low cost by making good use of BLE beacons and smartphone apps. In addition, the chapter focuses on efforts to collect the “emotions” of people who go and go on the street corner.},
 bibtype = {inbook},
 author = {Aihara, Kenro and Imura, Hajime},
 doi = {10.1007/978-3-030-15145-4_67-1},
 chapter = {Crowdsourcing for Smart Cities That Realizes the Situation of Cities and Information Sharing},
 title = {Handbook of Smart Cities}
}

The chapter describes initiatives aimed at improving the quality and efficiency of urban life by understanding the urban situation by utilizing an unspecified number of citizens and sharing collected and aggregated information, mainly using commoditized devices such as smartphones. Specifically, the chapter focuses on efforts to improve the efficiency of road management work by collecting road traffic conditions using a drive recorder application and acquiring sensor data that can estimate the road surface condition. It also shows a crowdsourced bus location service that can be realized at low cost by making good use of BLE beacons and smartphone apps. In addition, the chapter focuses on efforts to collect the “emotions” of people who go and go on the street corner.

@inproceedings{
 title = {Fully-Neural Approach to Heavy Vehicle Detection on Bridges Using a Single Strain Sensor},
 type = {inproceedings},
 year = {2020},
 pages = {3047-3051},
 month = {5},
 publisher = {IEEE},
 id = {977ab960-f787-3731-a638-74ba93f824f7},
 created = {2025-08-02T16:04:04.943Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:46:46.731Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:ICASSP:2020},
 private_publication = {false},
 abstract = {Bridge weigh-in-motion (BWIM) is a technique for detecting heavy vehicles that may cause serious damage to real bridges. BWIM is realized by analyzing the strain signals observed at places on the bridge in terms of bridge-component responses to the axle loads. In current practice, a BWIM system requires multiple strain sensors to collect vehicle properties including speed and axle positions for accurate load estimation, which may limit the system's life-span. Furthermore, BWIM should consider a wide variety of waveforms, which may be caused by vehicle acceleration and/or the various traveling positions in lanes. In this paper, we propose a novel BWIM mechanism, which employs a deep convolutional neural network (CNN). The CNN is able to learn actual traffic conditions and achieve accurate load estimation by using only a single strain sensor. The training dataset is collected from a distant load meter, by consulting traffic surveillance cameras and identifying similar vehicles. After the system initialization, the CNN requires no additional sensors (or cameras) for axle detection, which may reduce the costs of both installation and system maintenance.},
 bibtype = {inproceedings},
 author = {Kawakatsu, Takaya and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 doi = {10.1109/ICASSP40776.2020.9053137},
 booktitle = {ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}
}

Bridge weigh-in-motion (BWIM) is a technique for detecting heavy vehicles that may cause serious damage to real bridges. BWIM is realized by analyzing the strain signals observed at places on the bridge in terms of bridge-component responses to the axle loads. In current practice, a BWIM system requires multiple strain sensors to collect vehicle properties including speed and axle positions for accurate load estimation, which may limit the system's life-span. Furthermore, BWIM should consider a wide variety of waveforms, which may be caused by vehicle acceleration and/or the various traveling positions in lanes. In this paper, we propose a novel BWIM mechanism, which employs a deep convolutional neural network (CNN). The CNN is able to learn actual traffic conditions and achieve accurate load estimation by using only a single strain sensor. The training dataset is collected from a distant load meter, by consulting traffic surveillance cameras and identifying similar vehicles. After the system initialization, the CNN requires no additional sensors (or cameras) for axle detection, which may reduce the costs of both installation and system maintenance.

@inproceedings{
 title = {博物館における鑑賞体験の記念品化を目的とするナビゲーション端末操作ログからの印象深い展示物推定},
 type = {inproceedings},
 year = {2020},
 websites = {https://proceedings-of-deim.github.io/DEIM2020/papers/D4-5.pdf},
 month = {3},
 id = {cde0ae90-17a7-3978-86e3-f235c8269acc},
 created = {2025-08-02T16:05:48.228Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:17:31.660Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Takidaira:DEIM:2020},
 private_publication = {false},
 abstract = {本稿では，博物館内でのナビゲーションを行う iPad アプリケーションのログを分析することで，鑑賞体験そのものを記念品化したようなポストカードを生成するアルゴリズムを提案する．博物館や美術館において，インタラクティブなナビゲーション端末の登場に伴い，順路や見た展示物などの鑑賞体験が人によって異なる場合が増加している．現地の体験を思い出させ，記憶に定着させるきっかけのひとつに記念品があるが，ロゴや代表的展示物をあしらった従来の記念品では，このような個人化した鑑賞に対して十分な効果を発揮できない．そこで，ナビゲーション端末の操作ログを，印象や記憶に関する心理効果に基づいて分析することで，訪問者がどの展示物に強い印象を抱いたかを推定する．訪問者個人の印象に残った展示物を掲載したポストカードを記念品として配布することで，見返した際に鑑賞体験を強く思い返させ，印象をより深めさせたり，事後学習を促進できると考えられる．検証のため，国立民族学博物館の展示物データを用いて独自に開発したナビゲーション用 iPad アプリケーションである「みんぱくアプリ」のログから，実際にポストカードを生成する被験者実験を行った．「みんぱくに行く前日に下調べをする」というタスクにおける操作ログから，複数の検索モデルに基づくランキングを作成し，実際に記憶に残っているかをラベル付けした．これにより，博物館における印象深い展示物の推定に適したランキングモデルと，鑑賞体験を思い起こさせるのに有効な提示方法を明らかにした．},
 bibtype = {inproceedings},
 author = {瀧平, 士夫 and 莊司, 慶行 and 山本, 岳洋 and 山本, 祐輔 and 大島, 裕明 and 相原, 健郎 and 神門, 典子},
 booktitle = {第12回データ工学と情報マネジメントに関するフォーラム 論文集}
}

本稿では，博物館内でのナビゲーションを行う iPad アプリケーションのログを分析することで，鑑賞体験そのものを記念品化したようなポストカードを生成するアルゴリズムを提案する．博物館や美術館において，インタラクティブなナビゲーション端末の登場に伴い，順路や見た展示物などの鑑賞体験が人によって異なる場合が増加している．現地の体験を思い出させ，記憶に定着させるきっかけのひとつに記念品があるが，ロゴや代表的展示物をあしらった従来の記念品では，このような個人化した鑑賞に対して十分な効果を発揮できない．そこで，ナビゲーション端末の操作ログを，印象や記憶に関する心理効果に基づいて分析することで，訪問者がどの展示物に強い印象を抱いたかを推定する．訪問者個人の印象に残った展示物を掲載したポストカードを記念品として配布することで，見返した際に鑑賞体験を強く思い返させ，印象をより深めさせたり，事後学習を促進できると考えられる．検証のため，国立民族学博物館の展示物データを用いて独自に開発したナビゲーション用 iPad アプリケーションである「みんぱくアプリ」のログから，実際にポストカードを生成する被験者実験を行った．「みんぱくに行く前日に下調べをする」というタスクにおける操作ログから，複数の検索モデルに基づくランキングを作成し，実際に記憶に残っているかをラベル付けした．これにより，博物館における印象深い展示物の推定に適したランキングモデルと，鑑賞体験を思い起こさせるのに有効な提示方法を明らかにした．

@inproceedings{
 title = {個人の鑑賞軸獲得のためのミュージアム事前学習システム},
 type = {inproceedings},
 year = {2020},
 websites = {https://proceedings-of-deim.github.io/DEIM2020/papers/B4-4.pdf},
 month = {3},
 id = {1dd05be3-e9b4-3f30-b538-15914ab49bb6},
 created = {2025-08-02T16:07:28.984Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:18:49.895Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Yu:DEIM:2020},
 private_publication = {false},
 abstract = {本研究では，鑑賞者の興味や関心に基づいて，個人の鑑賞軸獲得のためのミュージアム事前学習システムを提案する．ミュージアムを訪れることは我々の生活の行動の中の一つである．しかし，ミュージアムを訪れた際，「学習目的を持ていない」，「時間が限られている」，「展示物が多すぎてを鑑賞しきれない」，「鑑賞に必要な知識が欠如している」といった理由で，ミュージアム疲れが起こる．ミュージアム疲れは，鑑賞した展示物が鑑賞者の記憶に残らない原因の一つであると考えられる．そのため，展示物を鑑賞者の記憶に残すには，鑑賞者が鑑賞軸を持つことが重要だと考える．以上のことから，本研究では，国立民族博物館を例として，鑑賞者が鑑賞軸を発見し，かつ，発見した鑑賞軸を反映した鑑賞コースマップを作成することができるミュージアム事前学習システムを提案する．},
 bibtype = {inproceedings},
 author = {楊, 澤華 and 莊司, 慶行 and 山本, 祐輔 and 山本, 岳洋 and 白石, 晃一 and 相原, 健郎 and 神門, 典子 and 大島, 裕明},
 booktitle = {第12回データ工学と情報マネジメントに関するフォーラム 論文集}
}

本研究では，鑑賞者の興味や関心に基づいて，個人の鑑賞軸獲得のためのミュージアム事前学習システムを提案する．ミュージアムを訪れることは我々の生活の行動の中の一つである．しかし，ミュージアムを訪れた際，「学習目的を持ていない」，「時間が限られている」，「展示物が多すぎてを鑑賞しきれない」，「鑑賞に必要な知識が欠如している」といった理由で，ミュージアム疲れが起こる．ミュージアム疲れは，鑑賞した展示物が鑑賞者の記憶に残らない原因の一つであると考えられる．そのため，展示物を鑑賞者の記憶に残すには，鑑賞者が鑑賞軸を持つことが重要だと考える．以上のことから，本研究では，国立民族博物館を例として，鑑賞者が鑑賞軸を発見し，かつ，発見した鑑賞軸を反映した鑑賞コースマップを作成することができるミュージアム事前学習システムを提案する．

@inproceedings{
 title = {On the Relationship between Accuracy of Bus Position Estimated by Crowdsourcing and Participation Density},
 type = {inproceedings},
 year = {2019},
 keywords = {Crowdsensing,Crowdsourcing,Internet of Things,Smart and hybrid cities},
 pages = {101-112},
 volume = {11587 LNCS},
 publisher = {Springer Verlag},
 id = {a30c9598-ff6b-3b0c-b8d7-865de73bd78d},
 created = {2021-10-12T05:24:02.051Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T17:07:39.501Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:HCII:2019},
 private_publication = {false},
 abstract = {The authors proposed a methodology of bus location service by crowdsource. In the conventional bus location service, a GPS receiver or the like for positioning the bus vehicle and a communication line for transmitting the position information are necessary for each bus and it causes expensive cost, whereas in this methodology, the BLE beacon and the application of the service installed in users’ smartphone transmits the beacon signal and the position information of the smartphone using the communication line of users’ smartphone so as to reduce the cost of bus operators. In order to grasp the position of each bus at any time on the server, one of the following is required: (1) At least one service user always exists for each bus, and (2) It is possible to estimate the position with high accuracy where there is no service user in a section. If contributing to the realization of smart city by grasping the situation in the town by crowdsourcing, it is extremely important to clarify the precision with respect to density that may be given as a relation between the number of users who participate in the program for grasping this situation and the size of the town. In this paper, the authors examine and show the relationship between the density of data collection and the accuracy of interpolation for the section where data is missing, using the position information actually obtained using the proposed application.},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Bin, Piao and Imura, Hajime},
 doi = {10.1007/978-3-030-21935-2_9},
 booktitle = {Distributed, Ambient and Pervasive Interactions, LNCS 11587}
}

The authors proposed a methodology of bus location service by crowdsource. In the conventional bus location service, a GPS receiver or the like for positioning the bus vehicle and a communication line for transmitting the position information are necessary for each bus and it causes expensive cost, whereas in this methodology, the BLE beacon and the application of the service installed in users’ smartphone transmits the beacon signal and the position information of the smartphone using the communication line of users’ smartphone so as to reduce the cost of bus operators. In order to grasp the position of each bus at any time on the server, one of the following is required: (1) At least one service user always exists for each bus, and (2) It is possible to estimate the position with high accuracy where there is no service user in a section. If contributing to the realization of smart city by grasping the situation in the town by crowdsourcing, it is extremely important to clarify the precision with respect to density that may be given as a relation between the number of users who participate in the program for grasping this situation and the size of the town. In this paper, the authors examine and show the relationship between the density of data collection and the accuracy of interpolation for the section where data is missing, using the position information actually obtained using the proposed application.

@article{
 title = {Deep Sensing Approach to Single-Sensor Vehicle Weighing System on Bridges},
 type = {article},
 year = {2019},
 keywords = {Deep neural network,bridge weigh-in-motion,structural health monitoring},
 pages = {253-256},
 volume = {19},
 month = {1},
 publisher = {Institute of Electrical and Electronics Engineers Inc.},
 day = {1},
 id = {00bb10d4-0974-36e5-9157-4abf800fd468},
 created = {2021-10-12T05:26:24.685Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T17:07:41.522Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Kawakatsu:IEEESensors:2019},
 private_publication = {false},
 abstract = {Bridge weigh-in-motion (BWIM) is a technique for detecting overloaded vehicles crossing a bridge without requiring them to stop. It may also be useful for monitoring the structural health of the bridge itself. To achieve accurate weighing of each vehicle, its properties, such as speed, locus, and wheel positions, should be estimated in advance. Conventionally, such information has been obtained via additional sensors such as cameras or via peak-signal detection, using multiple sensors installed across the bridge. This may require substantial computational resources or expensive synchronization between sensors, and the complexity of the overall BWIM system may lead to frequent breakdowns. In this paper, we propose a single-sensor-based BWIM system that utilizes a deep neural network. First, a vehicle's properties are obtained via feature extraction from the bridge strain response, as sampled by a single strain sensor. BWIM is then performed, using the same response data. The model parameters for vehicle detection are optimized automatically by consulting a surveillance camera while obtaining ground-Truth data for a large number of vehicles crossing the bridge. After the model is optimized for the target bridge, the camera may be removed. Our proposal paves the way toward low-cost, compact, and single-sensor BWIM systems.},
 bibtype = {article},
 author = {Kawakatsu, Takaya and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 doi = {10.1109/JSEN.2018.2872839},
 journal = {IEEE Sensors Journal},
 number = {1}
}

Bridge weigh-in-motion (BWIM) is a technique for detecting overloaded vehicles crossing a bridge without requiring them to stop. It may also be useful for monitoring the structural health of the bridge itself. To achieve accurate weighing of each vehicle, its properties, such as speed, locus, and wheel positions, should be estimated in advance. Conventionally, such information has been obtained via additional sensors such as cameras or via peak-signal detection, using multiple sensors installed across the bridge. This may require substantial computational resources or expensive synchronization between sensors, and the complexity of the overall BWIM system may lead to frequent breakdowns. In this paper, we propose a single-sensor-based BWIM system that utilizes a deep neural network. First, a vehicle's properties are obtained via feature extraction from the bridge strain response, as sampled by a single strain sensor. BWIM is then performed, using the same response data. The model parameters for vehicle detection are optimized automatically by consulting a surveillance camera while obtaining ground-Truth data for a large number of vehicles crossing the bridge. After the model is optimized for the target bridge, the camera may be removed. Our proposal paves the way toward low-cost, compact, and single-sensor BWIM systems.

@article{
 title = {サイバーフィジカルシステムでの実世界データ収集},
 type = {article},
 year = {2019},
 pages = {387-398},
 volume = {J102-B},
 id = {e0b9265b-846f-342f-b1fc-dc5bffcdbc1d},
 created = {2021-10-12T05:26:25.959Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:47:19.618Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:IEICE:2019},
 private_publication = {false},
 abstract = {無線通信環境の整備とデバイスの小型化・高機能化に加え，スマートフォンの普及やソーシャルネットワーキングサービス（SNS）の発展などを背景に，従来の情報学研究が主たる対象としてきたサイバー空間上のコンテンツでは捉えきれない実世界の状況把握と制御を目指したサイバーフィジカルシステム（CPS）への注目が広がってきている．CPSでは，物理世界の事象をセンシング技術等によりネットワーク越しに収集し，計算処理を施した結果を直接，あるいは間接に実世界にフィードバックし，その変化等をセンシングで捉えていくというサイクルを想定する．本論文では，CPSの概要について述べた後，主にセンシングからデータ収集の部分に着目する．ここでは，定点-移動体における計測と，センサーデバイスによる自動計測から利用者自身の投稿によるデータ収集までを整理した上で，三つの事例を中心に概説する．具体的には，インフラ構造物の長期モニタリング等を想定したセンシング，プローブカー等の移動体の計測と分析事例，及び，道路状況把握等のためのクラウドソーシングによるセンシングを取り上げる．これらの事例を通じて，今後の通信に求められる研究課題等を考えるきっかけを与えられることを期待する．},
 bibtype = {article},
 author = {相原, 健郎},
 doi = {10.14923/transcomj.2018MOI0002},
 journal = {電子情報通信学会論文誌B},
 number = {6}
}

無線通信環境の整備とデバイスの小型化・高機能化に加え，スマートフォンの普及やソーシャルネットワーキングサービス（SNS）の発展などを背景に，従来の情報学研究が主たる対象としてきたサイバー空間上のコンテンツでは捉えきれない実世界の状況把握と制御を目指したサイバーフィジカルシステム（CPS）への注目が広がってきている．CPSでは，物理世界の事象をセンシング技術等によりネットワーク越しに収集し，計算処理を施した結果を直接，あるいは間接に実世界にフィードバックし，その変化等をセンシングで捉えていくというサイクルを想定する．本論文では，CPSの概要について述べた後，主にセンシングからデータ収集の部分に着目する．ここでは，定点-移動体における計測と，センサーデバイスによる自動計測から利用者自身の投稿によるデータ収集までを整理した上で，三つの事例を中心に概説する．具体的には，インフラ構造物の長期モニタリング等を想定したセンシング，プローブカー等の移動体の計測と分析事例，及び，道路状況把握等のためのクラウドソーシングによるセンシングを取り上げる．これらの事例を通じて，今後の通信に求められる研究課題等を考えるきっかけを与えられることを期待する．

@inproceedings{
 title = {提示型検索モデルに基づくミュージアム鑑賞体験の提案},
 type = {inproceedings},
 year = {2019},
 month = {12},
 publisher = {情報処理学会},
 id = {58e3fbd5-341e-3deb-beb6-09923d54d35e},
 created = {2025-08-02T16:08:58.771Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:09:15.982Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kando:JinMonCom:2019},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {神門, 典子 and 大島, 裕明 and 相原, 健郎 and 莊司, 慶行 and 白石, 晃一 and 山本, 岳洋 and 山本, 祐輔 and 楊, 澤華},
 booktitle = {人文科学とコンピュータシンポジウム（じんもんこん2019） 論文集}
}

@inproceedings{
 title = {Development of Bridge Weigh-In-Motion Using Deck Slab Response},
 type = {inproceedings},
 year = {2019},
 pages = {1562-1567},
 month = {8},
 id = {f488425c-8820-3cf0-b743-a1cae3e73376},
 created = {2025-08-02T16:12:43.054Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:13:04.835Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Suzuki:ICSHMII:2019},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Suzuki, Keigo and Tamamura, Tomoya and Miyashita, Takeshi and Fukada, Saiji and Okano, Tsuyoshi and Aihara, Kenro and Takasu, Atsuhiro and Kinoshita, Akira},
 booktitle = {9th International Conference on Structural Health Monitoring of Intelligent Infrastructure}
}

@inproceedings{
 title = {Adversarial Media-fusion Approach to Strain Prediction for Bridges},
 type = {inproceedings},
 year = {2019},
 pages = {736-743},
 publisher = {SCITEPRESS - Science and Technology Publications},
 id = {45164629-2dde-3080-93ec-c481e4828c18},
 created = {2025-08-02T16:14:13.298Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:14:44.888Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:ICPRAM:2019},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Kawakatsu, Takaya and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 doi = {10.5220/0007555007360743},
 booktitle = {Proceedings of the 8th International Conference on Pattern Recognition Applications and Methods}
}

@inproceedings{
 title = {Collecting Bus Locations by Users: A Crowdsourcing Model to Estimate Operation Status of Bus Transit Service},
 type = {inproceedings},
 year = {2018},
 keywords = {Crowdsensing,Crowdsourcing,Internet of Things,Smart and hybrid cities},
 pages = {171-180},
 volume = {10921 LNCS},
 publisher = {Springer Verlag},
 id = {3574cdff-9be2-395c-b563-7dc56aca8869},
 created = {2021-10-12T05:24:00.742Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T17:07:46.012Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:HCII:2018},
 private_publication = {false},
 abstract = {This paper describes a crowdsourcing model to collect bus locations from onboard passengers. Bus location service, or realtime bus tracking service, is getting more and more general these days. Some models are proposed to build such services. One approach is facilitated every vehicle has a function to position itself with location sensor, such as GPS receiver, and transmits its own location with time to the server. Another is an environmental approach that bus detectors are deployed along the route to detect ids of nearby buses and transmit to the server. These models are well-established and practical. However, it is not easy to install such services especially for small operators because costs on devices and data transmission are relatively high. This paper proposes that a sustainable model even for small operators to provide bus locations to passengers. The key idea of the proposal is that collecting bus locations is not by bus operators but by onboard passengers. To collect them, a smartphone application of bus tracker is provided to public. The application shows current locations of buses in operation on bus transit services, while it detects nearby buses around users and transmits bus ids with time and location of detection to the service platform. That is, locations of buses are collected by users.},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Bin, Piao and Imura, Hajime and Takasu, Atsuhiro and Tanaka, Yuzuru},
 doi = {10.1007/978-3-319-91125-0_14},
 booktitle = {Distributed, Ambient and Pervasive Interactions, LNCS 10921}
}

This paper describes a crowdsourcing model to collect bus locations from onboard passengers. Bus location service, or realtime bus tracking service, is getting more and more general these days. Some models are proposed to build such services. One approach is facilitated every vehicle has a function to position itself with location sensor, such as GPS receiver, and transmits its own location with time to the server. Another is an environmental approach that bus detectors are deployed along the route to detect ids of nearby buses and transmit to the server. These models are well-established and practical. However, it is not easy to install such services especially for small operators because costs on devices and data transmission are relatively high. This paper proposes that a sustainable model even for small operators to provide bus locations to passengers. The key idea of the proposal is that collecting bus locations is not by bus operators but by onboard passengers. To collect them, a smartphone application of bus tracker is provided to public. The application shows current locations of buses in operation on bus transit services, while it detects nearby buses around users and transmits bus ids with time and location of detection to the service platform. That is, locations of buses are collected by users.

@article{
 title = {クラウドソースによるモバイルセンシングを用いた路面状況推定},
 type = {article},
 year = {2018},
 pages = {1784-1793},
 volume = {59},
 websites = {https://ipsj.ixsq.nii.ac.jp/records/191580},
 month = {10},
 id = {da08ee51-8f86-3708-96ec-086f3a4d2977},
 created = {2025-08-02T16:17:47.988Z},
 accessed = {2025-08-04},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:50:31.460Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Piao:IPSJ:2018},
 private_publication = {false},
 abstract = {路面状況の判別は，交通安全や運転の快適性を確保するために不可欠な技術である．本論文では，クラウドソースによるモバイルセンシングを用いた新たな路面状況推定手法を提案する．提案手法は，参加型センシングの観点から，一般の自動車利用者のスマートフォンをセンサとして，走行時に収集されたモーションセンサのみのデータを用いて路面状態（凍結，シャーベット，舗装路など）と路面形状（バンプ，穴，平坦など）の両方の推定を可能にする．提案手法では，2つのモジュール：1)車両走行モードクラスタリング（VTMC：Vehicle Travel Mode Clustering）モジュールと，2)路面状況分類（RSCC：Road Surface Condition Classification）モジュールを用いて路面の推定を行う．VTMCモジュールは，K-meansアルゴリズムを用いて各運転ごとの車両走行モードのクラスタリングすることで，ユーザ間の個人差を考慮する必要がなく，複数のユーザから収集したデータを統合した路面状況の推定を可能にする．一方RSCCモジュールは生成された各クラスタに振られ，Random Forest分類器を用いて各クラスタ内で路面状態の推定を行う．本論文では提案手法を実際のフィールドにおいて収集した走行データに適用し，その有用性についての検証を行った．その結果，提案手法による路面状況推定の平均正解率は90%を超える．},
 bibtype = {article},
 author = {朴, 斌 and 相原, 健郎},
 journal = {情報処理学会論文誌},
 number = {10}
}

路面状況の判別は，交通安全や運転の快適性を確保するために不可欠な技術である．本論文では，クラウドソースによるモバイルセンシングを用いた新たな路面状況推定手法を提案する．提案手法は，参加型センシングの観点から，一般の自動車利用者のスマートフォンをセンサとして，走行時に収集されたモーションセンサのみのデータを用いて路面状態（凍結，シャーベット，舗装路など）と路面形状（バンプ，穴，平坦など）の両方の推定を可能にする．提案手法では，2つのモジュール：1)車両走行モードクラスタリング（VTMC：Vehicle Travel Mode Clustering）モジュールと，2)路面状況分類（RSCC：Road Surface Condition Classification）モジュールを用いて路面の推定を行う．VTMCモジュールは，K-meansアルゴリズムを用いて各運転ごとの車両走行モードのクラスタリングすることで，ユーザ間の個人差を考慮する必要がなく，複数のユーザから収集したデータを統合した路面状況の推定を可能にする．一方RSCCモジュールは生成された各クラスタに振られ，Random Forest分類器を用いて各クラスタ内で路面状態の推定を行う．本論文では提案手法を実際のフィールドにおいて収集した走行データに適用し，その有用性についての検証を行った．その結果，提案手法による路面状況推定の平均正解率は90%を超える．

@inproceedings{
 title = {Deep Sensing Approach to Singel-Sensor Bridge Weighing in Motion},
 type = {inproceedings},
 year = {2018},
 id = {1ed7a059-f447-3987-8784-18f92fbe43a0},
 created = {2025-08-02T16:21:04.361Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:21:30.962Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:EWSHM:2018},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Kawakatsu, Takaya and Kinoshita, Akira and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 booktitle = {9th European Workshop on Structural Health Monitoring (EWSHM)}
}

@inbook{
 type = {inbook},
 year = {2018},
 pages = {49-58},
 volume = {LNCS 11031},
 publisher = {Springer},
 city = {Cham},
 id = {84917c42-b77a-3aac-8eca-c7957fa8f6f1},
 created = {2025-08-02T16:22:03.263Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:52:48.186Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:DaWaK:2018},
 private_publication = {false},
 abstract = {When a vehicle passes over a bridge, the bridge distorts in response to the vehicle’s load. The response characteristics may change over time if the bridge suffers damage. We consider the detection of such anomalous responses, using data from both traffic-surveillance cameras and strain sensors. The camera data are utilized to treat each vehicle’s identified properties as explanatory variables in the response model. The video and strain data are transformed into a common feature space, to enable direct comparisons. This space is obtained via our proposed spiral learning method, which is based on a deep convolutional neural network. We treat the distance between the video and strain data in the space as the anomaly score. We also propose an adversarial unsupervised learning technique for removing the influence of the weather. In our experiments, we found anomalous strain responses from a real bridge, and were able to classify them into three major patterns. The results demonstrate the effectiveness of our approach to bridge damage analysis.},
 bibtype = {inbook},
 author = {Kawakatsu, Takaya and Kinoshita, Akira and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 doi = {10.1007/978-3-319-98539-8_4},
 chapter = {Adversarial Spiral Learning Approach to Strain Analysis for Bridge Damage Detection},
 title = {Big Data Analytics and Knowledge Discovery. DaWaK 2018}
}

When a vehicle passes over a bridge, the bridge distorts in response to the vehicle’s load. The response characteristics may change over time if the bridge suffers damage. We consider the detection of such anomalous responses, using data from both traffic-surveillance cameras and strain sensors. The camera data are utilized to treat each vehicle’s identified properties as explanatory variables in the response model. The video and strain data are transformed into a common feature space, to enable direct comparisons. This space is obtained via our proposed spiral learning method, which is based on a deep convolutional neural network. We treat the distance between the video and strain data in the space as the anomaly score. We also propose an adversarial unsupervised learning technique for removing the influence of the weather. In our experiments, we found anomalous strain responses from a real bridge, and were able to classify them into three major patterns. The results demonstrate the effectiveness of our approach to bridge damage analysis.

@inproceedings{
 title = {Deep Sensing Approach to Vehicle Detection from Bridge Vibration},
 type = {inproceedings},
 year = {2018},
 id = {5c23c84f-6f40-311d-b8ad-976a8dc84f5e},
 created = {2025-08-02T16:24:13.322Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:24:30.143Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:ISIP:2018},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Kawakatsu, Takaya and Kinoshita, Akira and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 booktitle = {12th International Workshop on Information Search, Integration, and Personalization (ISIP)}
}

@inproceedings{
 title = {A Smart City Application for Sharing Up-to-date Road Surface Conditions Detected from Crowdsourced Data},
 type = {inproceedings},
 year = {2017},
 keywords = {City,Crowdsourcing,Cyber-physical systems,Mobile sensing,Road condition detection,Smart},
 pages = {219-234},
 volume = {10291 LNCS},
 id = {1fe6554c-08b1-3230-94cb-5571b737064b},
 created = {2017-08-07T22:13:48.847Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:31:26.843Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:HCII:2017},
 private_publication = {false},
 abstract = {© Springer International Publishing AG 2017. This paper introduces a smart city application to share road conditions. The application is based on a mobile sensing framework to collect sensor data reflecting personal-scale, or microscopic, roadside phenomena using crowdsourcing. To collect data, a driving recorder smartphone application that records not only sensor data but also videos from the driver’s view is used. To extract specific roadside phenomena, collected data are integrated and analyzed at the service platform. One example is estimating road surface conditions. The paper shows our method to estimate road surface type (RST) and road surface shape (RSS). Features are defined in Sequential Forward Floating Search (SFFS) algorithm from collected data. By using random forest as classifier, average recall was about 91% in the 50 km/h –80km/h range. The result may support to build a service that provides detected road conditions from up-to-date crowdsourced mobile sensing application.},
 bibtype = {inproceedings},
 author = {Aihara, K. and Bin, P. and Imura, H. and Takasu, A. and Tanaka, Y.},
 doi = {10.1007/978-3-319-58697-7_16},
 booktitle = {Distributed, Ambient and Pervasive Interactions, LNCS 10291}
}

© Springer International Publishing AG 2017. This paper introduces a smart city application to share road conditions. The application is based on a mobile sensing framework to collect sensor data reflecting personal-scale, or microscopic, roadside phenomena using crowdsourcing. To collect data, a driving recorder smartphone application that records not only sensor data but also videos from the driver’s view is used. To extract specific roadside phenomena, collected data are integrated and analyzed at the service platform. One example is estimating road surface conditions. The paper shows our method to estimate road surface type (RST) and road surface shape (RSS). Features are defined in Sequential Forward Floating Search (SFFS) algorithm from collected data. By using random forest as classifier, average recall was about 91% in the 50 km/h –80km/h range. The result may support to build a service that provides detected road conditions from up-to-date crowdsourced mobile sensing application.

@inproceedings{
 title = {Collaborative Item Embedding Model for Implicit Feedback Data},
 type = {inproceedings},
 year = {2017},
 keywords = {Collaborative filtering,Item embedding,Matrix factorization,Recommender system},
 pages = {336-348},
 volume = {10360 LNCS},
 id = {fc79b1e3-a1e4-3884-b41f-a5b143c5c271},
 created = {2017-12-20T04:28:44.140Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T17:07:53.386Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {© Springer International Publishing AG 2017. Collaborative filtering is the most popular approach for recommender systems. One way to perform collaborative filtering is matrix factorization, which characterizes user preferences and item attributes using latent vectors. These latent vectors are good at capturing global features of users and items but are not strong in capturing local relationships between users or between items. In this work, we propose a method to extract the relationships between items and embed them into the latent vectors of the factorization model. This combines two worlds: matrix factorization for collaborative filtering and item embedding, a similar concept to word embedding in language processing. Our experiments on three real-world datasets show that our proposed method outperforms competing methods on top-n recommendation tasks.},
 bibtype = {inproceedings},
 author = {Nguyen, ThaiBinh and Aihara, Kenro and Takasu, Atsuhiro},
 doi = {10.1007/978-3-319-60131-1_19},
 booktitle = {International Conference on Web Engineering, LNCS 10360}
}

© Springer International Publishing AG 2017. Collaborative filtering is the most popular approach for recommender systems. One way to perform collaborative filtering is matrix factorization, which characterizes user preferences and item attributes using latent vectors. These latent vectors are good at capturing global features of users and items but are not strong in capturing local relationships between users or between items. In this work, we propose a method to extract the relationships between items and embed them into the latent vectors of the factorization model. This combines two worlds: matrix factorization for collaborative filtering and item embedding, a similar concept to word embedding in language processing. Our experiments on three real-world datasets show that our proposed method outperforms competing methods on top-n recommendation tasks.

@inproceedings{
 title = {Traffic Surveillance System for Bridge Vibration Analysis},
 type = {inproceedings},
 year = {2017},
 pages = {69-74},
 volume = {2017-Janua},
 id = {a611968a-b94b-3f98-b2f3-7c7eacf3f0ca},
 created = {2018-04-04T10:49:22.058Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:24:59.474Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kawakatsu:IRI:2017},
 private_publication = {false},
 abstract = {© 2017 IEEE. The vibration response of a damaged bridge is known to have changed characteristics. To analyze the response, we start by collecting waveforms of the vibration immediately following the passage of a vehicle. We then need to isolate just those vibrations caused by a single heavy vehicle, if the vibration characteristics are to be accurate. In this paper, we propose a traffic-vibration analysis system that interacts with a surveillance camera. The system identifies a vehicle from the video by combining a moving-object detector with a neural-network-based object detector, thereby estimating automatically the bridge's natural frequencies and damping ratios as features that characterize the bridge's damage.},
 bibtype = {inproceedings},
 author = {Kawakatsu, Takaya and Kakitani, Akira and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 doi = {10.1109/IRI.2017.82},
 booktitle = {2017 IEEE International Conference on Information Reuse and Integration, IRI 2017}
}

© 2017 IEEE. The vibration response of a damaged bridge is known to have changed characteristics. To analyze the response, we start by collecting waveforms of the vibration immediately following the passage of a vehicle. We then need to isolate just those vibrations caused by a single heavy vehicle, if the vibration characteristics are to be accurate. In this paper, we propose a traffic-vibration analysis system that interacts with a surveillance camera. The system identifies a vehicle from the video by combining a moving-object detector with a neural-network-based object detector, thereby estimating automatically the bridge's natural frequencies and damping ratios as features that characterize the bridge's damage.

@inbook{
 type = {inbook},
 year = {2017},
 pages = {271-297},
 month = {5},
 publisher = {Springer International Publishing},
 day = {29},
 id = {95a96f87-2a13-3858-9970-3e8237bec2c2},
 created = {2021-10-12T05:23:57.028Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:54:49.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:SmartSensors:2017},
 private_publication = {false},
 abstract = {This paper describes a proposed mobile sensing framework for collecting sensor data reflecting personal-scale, or microscopic, roadside phenomena through crowdsourcing and the use of big data, such as traffic and climate data, as well as the contents of social networking services such as Twitter. To collect these data, a smartphone application is provided. One feature is a driving recorder that collects not only sensor data but also videos recorded from the driver’s point of view. To extract specific roadside phenomena, the collected data are integrated and analyzed using the service platform. The proposed smartphone application offers the following advantages over appliances: (1) ordinary appliances are stand-alone, which means that storage is limited, whereas the application is connected to the cloud; (2) appliances are not cheap, whereas the application is free; and (3) appliances only store driving records, whereas the application can obtain feedback from the service platform. The authors expect that these advantages will provide citizens with the necessary incentive to use the application. We conducted an experimental survey using our prototype application to examine how effective such functionality is in motivating users. The results showed that most users found the application useful.},
 bibtype = {inbook},
 author = {Aihara, Kenro and Imura, Hajime and Piao, Bin and Takasu, Atsuhiro and Tanaka, Yuzuru},
 doi = {10.1007/978-3-319-55345-0_11},
 chapter = {Mobile Crowdsensing to Collect Road Conditions and Events},
 title = {Smart Sensors at the IoT Frontier}
}

This paper describes a proposed mobile sensing framework for collecting sensor data reflecting personal-scale, or microscopic, roadside phenomena through crowdsourcing and the use of big data, such as traffic and climate data, as well as the contents of social networking services such as Twitter. To collect these data, a smartphone application is provided. One feature is a driving recorder that collects not only sensor data but also videos recorded from the driver’s point of view. To extract specific roadside phenomena, the collected data are integrated and analyzed using the service platform. The proposed smartphone application offers the following advantages over appliances: (1) ordinary appliances are stand-alone, which means that storage is limited, whereas the application is connected to the cloud; (2) appliances are not cheap, whereas the application is free; and (3) appliances only store driving records, whereas the application can obtain feedback from the service platform. The authors expect that these advantages will provide citizens with the necessary incentive to use the application. We conducted an experimental survey using our prototype application to examine how effective such functionality is in motivating users. The results showed that most users found the application useful.

@inproceedings{
 title = {Detecting the road surface condition by using mobile crowdsensing with drive recorder},
 type = {inproceedings},
 year = {2017},
 pages = {1-8},
 month = {10},
 publisher = {IEEE},
 id = {4b465bb8-1622-3314-a412-9dfc24a5c071},
 created = {2025-08-02T16:19:03.082Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:55:09.308Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Piao:IEEEITS:2017},
 private_publication = {false},
 abstract = {Road surface conditions affect transport safety and driving comfort, particularly in snowy areas. This paper proposes a new method for detecting road surface conditions based on crowdsourced mobile sensing technology. The method can efficiently detect road surface conditions using motion sensors embedded in smartphones mounted on vehicles. Detecting road conditions using such sensors, which are usually loosely placed in the vehicle, nonetheless poses a challenge. Our approach comprises two modules: a clustering module and classification module. The clustering module uses a K-means algorithm to distribute the vehicle travel mode into a proper cluster. The classification module uses the random forest classifier, which is assigned to each group of vehicle travel mode, to classify the road surface conditions. We defined new road surface conditions as the estimation target, considering both the substance that covers the road surface and the shape of the road surface itself. The results show that our approach can detect road surface conditions with accuracy as high as 90%.},
 bibtype = {inproceedings},
 author = {Piao, Bin and Aihara, Kenro},
 doi = {10.1109/ITSC.2017.8317818},
 booktitle = {2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC)}
}

Road surface conditions affect transport safety and driving comfort, particularly in snowy areas. This paper proposes a new method for detecting road surface conditions based on crowdsourced mobile sensing technology. The method can efficiently detect road surface conditions using motion sensors embedded in smartphones mounted on vehicles. Detecting road conditions using such sensors, which are usually loosely placed in the vehicle, nonetheless poses a challenge. Our approach comprises two modules: a clustering module and classification module. The clustering module uses a K-means algorithm to distribute the vehicle travel mode into a proper cluster. The classification module uses the random forest classifier, which is assigned to each group of vehicle travel mode, to classify the road surface conditions. We defined new road surface conditions as the estimation target, considering both the substance that covers the road surface and the shape of the road surface itself. The results show that our approach can detect road surface conditions with accuracy as high as 90%.

@article{
 title = {グループ会話における発話意図の推定システム},
 type = {article},
 year = {2017},
 pages = {1-11},
 volume = {58},
 websites = {https://ipsj.ixsq.nii.ac.jp/records/178936},
 month = {5},
 id = {0099ddfa-ca50-3085-8ea5-fa209503b4db},
 created = {2025-08-02T16:26:31.659Z},
 accessed = {2025-08-04},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:35:30.201Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Kono:IPSJ:2017},
 private_publication = {false},
 abstract = {一緒に旅行に出かけるグループは，旅行を楽しむといった共通目標を持つものの，その目標の重要性や旅行中に達成したいことに対する各メンバの意識が違っている場合があり，決議の結果でその意思決定を評価することが必ずしも適さず，意思決定のプロセスが重要なことがある．このようなカジュアルなグループ意思決定において，意見を述べられていないメンバが意見を述べられるようになり，また，各メンバの意見について討議されるようになれば，充実したコミュニケーションを通じてグループ内の良好な関係維持につながると想定される．このため，筆者らは会話状況等を含むグループのコンテキストを推定し，そのグループ状態に応じた適切なアドバイスを会話エージェントを通じて提示するといった介入によるグループ会話における意思決定の支援を検討している．このようなグループにおける意思決定の支援を，会話分析を通じて実現するには，会話参加者のコンテキストを獲得する必要がある．しかしながら，雑音等が多く発生するような車内環境での音声認識精度は実用上は不十分であり，また，正確に音声を認識できた場合でも語彙が少ない発話等では意図等の抽出が困難な場合があり，音声認識を利用した発話意図等の推定では部分的に欠落を生じる課題がある．本論文では，上記の課題を解決するため，発話テキストから意図を推定するとともに，テキストから意図が推定できなかった部分については，発話の音響的特徴量データを利用し意図を推定して，会話内の各発話における意図を特定するシステムを提案する．テスト会話による実験の結果，音声認識による発話テキストのみを用いて発話意図を推定する従来手法よりも，音声認識による発話テキストに加え音響的特徴量データを併用して発話意図を推定する提案手法の方が，高い精度で発話意図を推定できることが示された．

In this paper, the authors propose a system of extracting utterance intentions from conversation in mobile environment, such as car. The authors assume that a group decision support in such environment, it is not enough to evaluate the result of decision but to evaluate the quality of its process, such as individual satisfaction of the discussion. To enhance the overall condition of the group, the authors have been considered a supporting environment to provide appropriate reference information in a timely manner through a conversational agent. For realizing this group decision support, the authors propose a system of estimating utterance intentions by using not only the converted text data but also voice feature values of each utterance in a conversation. As a result of experiments, almost all of the intentions of utterances in a conversation can be estimated by this proposed system, even in the cases of the voice recognition or intention extraction failed.},
 bibtype = {article},
 author = {河野, 進 and 相原, 健郎},
 journal = {情報処理学会論文誌},
 number = {5}
}

一緒に旅行に出かけるグループは，旅行を楽しむといった共通目標を持つものの，その目標の重要性や旅行中に達成したいことに対する各メンバの意識が違っている場合があり，決議の結果でその意思決定を評価することが必ずしも適さず，意思決定のプロセスが重要なことがある．このようなカジュアルなグループ意思決定において，意見を述べられていないメンバが意見を述べられるようになり，また，各メンバの意見について討議されるようになれば，充実したコミュニケーションを通じてグループ内の良好な関係維持につながると想定される．このため，筆者らは会話状況等を含むグループのコンテキストを推定し，そのグループ状態に応じた適切なアドバイスを会話エージェントを通じて提示するといった介入によるグループ会話における意思決定の支援を検討している．このようなグループにおける意思決定の支援を，会話分析を通じて実現するには，会話参加者のコンテキストを獲得する必要がある．しかしながら，雑音等が多く発生するような車内環境での音声認識精度は実用上は不十分であり，また，正確に音声を認識できた場合でも語彙が少ない発話等では意図等の抽出が困難な場合があり，音声認識を利用した発話意図等の推定では部分的に欠落を生じる課題がある．本論文では，上記の課題を解決するため，発話テキストから意図を推定するとともに，テキストから意図が推定できなかった部分については，発話の音響的特徴量データを利用し意図を推定して，会話内の各発話における意図を特定するシステムを提案する．テスト会話による実験の結果，音声認識による発話テキストのみを用いて発話意図を推定する従来手法よりも，音声認識による発話テキストに加え音響的特徴量データを併用して発話意図を推定する提案手法の方が，高い精度で発話意図を推定できることが示された． In this paper, the authors propose a system of extracting utterance intentions from conversation in mobile environment, such as car. The authors assume that a group decision support in such environment, it is not enough to evaluate the result of decision but to evaluate the quality of its process, such as individual satisfaction of the discussion. To enhance the overall condition of the group, the authors have been considered a supporting environment to provide appropriate reference information in a timely manner through a conversational agent. For realizing this group decision support, the authors propose a system of estimating utterance intentions by using not only the converted text data but also voice feature values of each utterance in a conversation. As a result of experiments, almost all of the intentions of utterances in a conversation can be estimated by this proposed system, even in the cases of the voice recognition or intention extraction failed.

@article{
 title = {ビッグデータを用いた観光動態把握とその活用 〜 動体データで訪日外客の動きを捉える},
 type = {article},
 year = {2017},
 pages = {743-754},
 month = {2},
 id = {7a78cd49-b509-3628-a391-6ec806793b47},
 created = {2025-08-02T16:27:48.253Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:58:40.795Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:JohoKanri:2017},
 private_publication = {false},
 abstract = {本稿は，携帯電話や位置に基づくサービス等で取得される動体データを用いて，その解析に基づく観光動態把握と活用に関して，分析事例なども交えて概説する。まず，観光分析の観点を整理したうえで，携帯電話基地局における動体データ，スマートフォンアプリケーションを用いた動体データ，および，SNSデータ等の性質などを紹介する。次に，観光庁の訪日外客の動態調査で行われた分析事例を紹介する。加えて，分析における問題も指摘する。さらに，目的地性指標を用いた動体データの分析結果を示し，人間の行動モデルの一つであるLevy Flightモデルと一致することを紹介する。},
 bibtype = {article},
 author = {相原, 健郎},
 journal = {情報管理}
}

本稿は，携帯電話や位置に基づくサービス等で取得される動体データを用いて，その解析に基づく観光動態把握と活用に関して，分析事例なども交えて概説する。まず，観光分析の観点を整理したうえで，携帯電話基地局における動体データ，スマートフォンアプリケーションを用いた動体データ，および，SNSデータ等の性質などを紹介する。次に，観光庁の訪日外客の動態調査で行われた分析事例を紹介する。加えて，分析における問題も指摘する。さらに，目的地性指標を用いた動体データの分析結果を示し，人間の行動モデルの一つであるLevy Flightモデルと一致することを紹介する。

@inbook{
 type = {inbook},
 year = {2017},
 pages = {336-348},
 id = {7081b615-4b41-38b4-83ce-ca2fb4cbc034},
 created = {2025-08-02T16:28:28.925Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:45:23.628Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Nguyen:LNCS:2017},
 private_publication = {false},
 abstract = {Collaborative filtering is the most popular approach for recommender systems. One way to perform collaborative filtering is matrix factorization, which characterizes user preferences and item attributes using latent vectors. These latent vectors are good at capturing global features of users and items but are not strong in capturing local relationships between users or between items. In this work, we propose a method to extract the relationships between items and embed them into the latent vectors of the factorization model. This combines two worlds: matrix factorization for collaborative filtering and item embedding, a similar concept to word embedding in language processing. Our experiments on three real-world datasets show that our proposed method outperforms competing methods on top-n recommendation tasks.},
 bibtype = {inbook},
 author = {Nguyen, ThaiBinh and Aihara, Kenro and Takasu, Atsuhiro},
 doi = {10.1007/978-3-319-60131-1_19},
 chapter = {Collaborative Item Embedding Model for Implicit Feedback Data}
}

Collaborative filtering is the most popular approach for recommender systems. One way to perform collaborative filtering is matrix factorization, which characterizes user preferences and item attributes using latent vectors. These latent vectors are good at capturing global features of users and items but are not strong in capturing local relationships between users or between items. In this work, we propose a method to extract the relationships between items and embed them into the latent vectors of the factorization model. This combines two worlds: matrix factorization for collaborative filtering and item embedding, a similar concept to word embedding in language processing. Our experiments on three real-world datasets show that our proposed method outperforms competing methods on top-n recommendation tasks.

@inbook{
 type = {inbook},
 year = {2017},
 pages = {66-81},
 id = {9ec5de6e-2c0f-3294-b2a1-9e7e030853e5},
 created = {2025-08-02T16:29:20.195Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:45:42.765Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Piao:CCIS:2017},
 private_publication = {false},
 abstract = {Road surface conditions have a significant impact on transport safety and driving comfort, particularly in snowy areas. This paper proposes a new method for estimating road surface conditions by using a motion sensor embedded in a smartphone. The method is based on a mobile sensing framework that can collect sensor data using crowdsourcing. In this study, we have defined new road surface conditions as the estimation target, which takes into account both the substance that covers the road surface, and the shape of the road surface itself. The paper also describes a method of feature selection, comprising two steps: First, an initial feature set is directly calculated using various features published in previous studies, using raw sensor data. Second, three feature selection algorithms are compared: Principal Component Analysis (PCA), Relief-F, and Sequential Forward Floating Search (SFFS), and the most effective of the three chosen. In this study, the SFFS algorithm showed higher accuracy than the others. The road surface condition classification was performed across different speed ranges using the Random Forest Classifier, and results show that the best accuracy, of about 91%, was obtained in the 50 km/h–80 km/h range.},
 bibtype = {inbook},
 author = {Piao, Bin and Aihara, Kenro and Kinoshita, Akira and Takasu, Atsuhiro and Adachi, Jun},
 doi = {10.1007/978-3-319-68282-2_5},
 chapter = {Estimating Road Surface Condition Using Crowdsourcing}
}

Road surface conditions have a significant impact on transport safety and driving comfort, particularly in snowy areas. This paper proposes a new method for estimating road surface conditions by using a motion sensor embedded in a smartphone. The method is based on a mobile sensing framework that can collect sensor data using crowdsourcing. In this study, we have defined new road surface conditions as the estimation target, which takes into account both the substance that covers the road surface, and the shape of the road surface itself. The paper also describes a method of feature selection, comprising two steps: First, an initial feature set is directly calculated using various features published in previous studies, using raw sensor data. Second, three feature selection algorithms are compared: Principal Component Analysis (PCA), Relief-F, and Sequential Forward Floating Search (SFFS), and the most effective of the three chosen. In this study, the SFFS algorithm showed higher accuracy than the others. The road surface condition classification was performed across different speed ranges using the Random Forest Classifier, and results show that the best accuracy, of about 91%, was obtained in the 50 km/h–80 km/h range.

@book{
 title = {A model of improving individual satisfaction in group conversation based on estimation of group status},
 type = {book},
 year = {2016},
 source = {Advances in Intelligent Systems and Computing},
 keywords = {[Conversation estimation, Conversational agent, Gr},
 volume = {487},
 id = {b977bac3-cc7b-3655-9351-503c59466847},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:32:49.970Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kono:ASOE:2017},
 private_publication = {false},
 abstract = {© Springer International Publishing Switzerland 2017.We generated a model to improve the satisfaction individual group members involved in a decision-making conversation by the intervention such as eliciting opinions from the members that cannot express in the discussion for group decision. We assume to estimate the group status by measuring the utterance (phonetic) characteristics in the conversation. Our goal is to develop a model that would provide decision support. In group decision making, some members may hesitate or yield their position to superiors, or they may offer no concrete ideas or suggestions, and thus may feel frustration during the conversation. Members who are in a weak position may not be as confident and may not be able to clearly express their intentions. As a test case, we consider the challenge of providing support to a group during a decision-making process (e.g., route, stopover) while traveling in a car. We sought to measure the satisfaction of each member during the process of decision making. Furthermore, the relationship between individual satisfaction and receptivity depended on how a member was accepted during the group discussions. We found that several participants in the test conversations exhibited increasing levels of stress when proposals were rejected or not thoroughly discussed. Thus, we think that reducing the frequency of such situations may be relevant in improving decision-making satisfaction. Therefore, we propose that a model in which all members of the group are given an equal opportunity to make proposals, and in which no members face continuous rejection during the decision-making conversation, will produce the greatest satisfaction. We plan to continue to study this model to facilitate the design of a decision support system.},
 bibtype = {book},
 author = {Kono, S. and Aihara, K.},
 doi = {10.1007/978-3-319-41688-5_22}
}

© Springer International Publishing Switzerland 2017.We generated a model to improve the satisfaction individual group members involved in a decision-making conversation by the intervention such as eliciting opinions from the members that cannot express in the discussion for group decision. We assume to estimate the group status by measuring the utterance (phonetic) characteristics in the conversation. Our goal is to develop a model that would provide decision support. In group decision making, some members may hesitate or yield their position to superiors, or they may offer no concrete ideas or suggestions, and thus may feel frustration during the conversation. Members who are in a weak position may not be as confident and may not be able to clearly express their intentions. As a test case, we consider the challenge of providing support to a group during a decision-making process (e.g., route, stopover) while traveling in a car. We sought to measure the satisfaction of each member during the process of decision making. Furthermore, the relationship between individual satisfaction and receptivity depended on how a member was accepted during the group discussions. We found that several participants in the test conversations exhibited increasing levels of stress when proposals were rejected or not thoroughly discussed. Thus, we think that reducing the frequency of such situations may be relevant in improving decision-making satisfaction. Therefore, we propose that a model in which all members of the group are given an equal opportunity to make proposals, and in which no members face continuous rejection during the decision-making conversation, will produce the greatest satisfaction. We plan to continue to study this model to facilitate the design of a decision support system.

@book{
 title = {Prototype of decision support based on estimation of group status using conversation analysis},
 type = {book},
 year = {2016},
 source = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
 keywords = {[Conversation estimation, Conversational agent, Gr},
 volume = {9735},
 id = {bcbe2f12-8717-36c3-abff-404fb1fef15c},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:33:56.548Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kono:HIMIAS:2016},
 private_publication = {false},
 abstract = {© Springer International Publishing Switzerland 2016.We propose a prototype for a system to verify the decision support of a group based on estimations of group status through utterance analysis. Based on methods used in prior studies of group dynamics and utterance analysis, we measured the utterance characteristics of group members to infer group status; moreover, we aimed to enhance the overall condition of the group by providing appropriate reference information in a timely manner through a conversational agent system. The goal of this system was a more satisfying decision-making process. We manufactured a prototype system to verify both the operations involved in the test case and the ability to infer group classification and status according to group dynamics. Our proposed method was appropriate based on the prototype system. Future work will focus on optimizing the logic and system functions of group status estimation and the subsequent step of informing the group.},
 bibtype = {book},
 author = {Kono, S. and Aihara, K.},
 doi = {10.1007/978-3-319-40397-7_5}
}

© Springer International Publishing Switzerland 2016.We propose a prototype for a system to verify the decision support of a group based on estimations of group status through utterance analysis. Based on methods used in prior studies of group dynamics and utterance analysis, we measured the utterance characteristics of group members to infer group status; moreover, we aimed to enhance the overall condition of the group by providing appropriate reference information in a timely manner through a conversational agent system. The goal of this system was a more satisfying decision-making process. We manufactured a prototype system to verify both the operations involved in the test case and the ability to infer group classification and status according to group dynamics. Our proposed method was appropriate based on the prototype system. Future work will focus on optimizing the logic and system functions of group status estimation and the subsequent step of informing the group.

@book{
 title = {On feasibility of crowdsourced mobile sensing for smarter city life},
 type = {book},
 year = {2016},
 source = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
 volume = {9749},
 id = {816c8916-12db-3563-9a58-2c05f2dc2042},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:33:08.370Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:HCII:2016},
 private_publication = {false},
 abstract = {© Springer International Publishing Switzerland 2016.This paper introduces the ongoing project that aims to develop a mobile sensing framework to collect sensor data reflecting personal-scale, or microscopic, roadside phenomena by crowdsourcing and also using social big data, such as traffic, climate, and contents of social network services like Twitter. To collect them, smartphone applications are provided. One of the typical applications is a driving recorder that collects not only sensor data but also recorded videos from the driver’s view. To extract specific roadside phenomena, collected data are integrated and analyzed at the service platform. The proposed smartphone application can be replaced with appliances because of its advantages: (1) ordinary appliances work stand-alone, which means that local storage is limited; the application is connected to the cloud, (2) appliances are not cheep, at least users must pay for it; the application is free, (3) appliances only store driving records; the application can get feedback from the service. The authors expect that these advantages can be accepted by citizen as an incentive to use it. To reveal how effective such function is for users’ motivation, an experiment and a survey are conducted with our prototyped service. As a result, most of the users accepted the function as attractive to use.},
 bibtype = {book},
 author = {Aihara, K. and Bin, P. and Imura, H. and Takasu, A. and Tanaka, Y.},
 doi = {10.1007/978-3-319-39862-4_36}
}

© Springer International Publishing Switzerland 2016.This paper introduces the ongoing project that aims to develop a mobile sensing framework to collect sensor data reflecting personal-scale, or microscopic, roadside phenomena by crowdsourcing and also using social big data, such as traffic, climate, and contents of social network services like Twitter. To collect them, smartphone applications are provided. One of the typical applications is a driving recorder that collects not only sensor data but also recorded videos from the driver’s view. To extract specific roadside phenomena, collected data are integrated and analyzed at the service platform. The proposed smartphone application can be replaced with appliances because of its advantages: (1) ordinary appliances work stand-alone, which means that local storage is limited; the application is connected to the cloud, (2) appliances are not cheep, at least users must pay for it; the application is free, (3) appliances only store driving records; the application can get feedback from the service. The authors expect that these advantages can be accepted by citizen as an incentive to use it. To reveal how effective such function is for users’ motivation, an experiment and a survey are conducted with our prototyped service. As a result, most of the users accepted the function as attractive to use.

@inproceedings{
 title = {GPS trajectory data enrichment based on a latent statistical model},
 type = {inproceedings},
 year = {2016},
 keywords = {[GPS trajectory data, Interpolation, Latent statis},
 id = {54a5dace-e4ed-3215-a29f-25ef239b0528},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:38:37.249Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kinoshita:ICPRAM:2016},
 private_publication = {false},
 abstract = {© Copyright 2016 by SCITEPRESS - Science and Technology Publications, Lda. All rights reserved.This paper proposes a latent statistical model for analyzing global positioning system (GPS) trajectory data. Because of the rapid spread of GPS-equipped devices, numerous GPS trajectories have become available, and they are useful for various location-aware systems. To better utilize GPS data, a number of sensor data mining techniques have been developed. This paper discusses the application of a latent statistical model to two closely related problems, namely, moving mode estimation and interpolation of the GPS observation. The proposed model estimates a latent mode of moving objects and represents moving patterns according to the mode by exploiting a large GPS trajectory dataset. We evaluate the effectiveness of the model through experiments using the GeoLife GPS Trajectories dataset and show that more than three-quarters of covered locations were correctly reproduced by interpolation at a fine granularity.},
 bibtype = {inproceedings},
 author = {Kinoshita, A. and Takasu, A. and Aihara, K. and Ishii, J. and Kurasawa, H. and Sato, H. and Nakamura, M. and Adachi, J.},
 booktitle = {ICPRAM 2016 - Proceedings of the 5th International Conference on Pattern Recognition Applications and Methods}
}

© Copyright 2016 by SCITEPRESS - Science and Technology Publications, Lda. All rights reserved.This paper proposes a latent statistical model for analyzing global positioning system (GPS) trajectory data. Because of the rapid spread of GPS-equipped devices, numerous GPS trajectories have become available, and they are useful for various location-aware systems. To better utilize GPS data, a number of sensor data mining techniques have been developed. This paper discusses the application of a latent statistical model to two closely related problems, namely, moving mode estimation and interpolation of the GPS observation. The proposed model estimates a latent mode of moving objects and represents moving patterns according to the mode by exploiting a large GPS trajectory dataset. We evaluate the effectiveness of the model through experiments using the GeoLife GPS Trajectories dataset and show that more than three-quarters of covered locations were correctly reproduced by interpolation at a fine granularity.

@inproceedings{
 title = {移動手段の混在を考慮した移動経路推定法},
 type = {inproceedings},
 year = {2016},
 month = {3},
 publisher = {電子情報通信学会},
 id = {760ae945-f294-3e5b-ba45-1f943c1671b3},
 created = {2025-08-02T16:35:26.349Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:35:44.906Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kurasawa:IEICE:2016},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {倉沢, 央 and 佐藤, 浩史 and 中村, 元紀 and 木下, 僚 and 高須, 淳宏 and 相原, 健郎 and 安達, 淳},
 booktitle = {2016年度電子情報通信学会総合大会講演論文集}
}

@inproceedings{
 title = {実世界における行動の選択と誘引},
 type = {inproceedings},
 year = {2016},
 month = {3},
 publisher = {電子情報通信学会},
 id = {6b344451-849b-3218-a982-a9092cda8e80},
 created = {2025-08-02T16:36:48.706Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:36:58.533Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:IEICE:2016},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎},
 booktitle = {2016年度電子情報通信学会総合大会講演論文集}
}

@inproceedings{
 title = {意思決定会話における満足度についての考察},
 type = {inproceedings},
 year = {2016},
 pages = {1-4},
 month = {3},
 publisher = {電子情報通信学会},
 id = {a677275d-2c93-3252-8f25-d362ed5c3ea7},
 created = {2025-08-02T16:37:57.774Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:38:14.975Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kono:IEICE:2016},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {河野, 進 and 相原, 健郎},
 booktitle = {信学技報}
}

@book{
 title = {A model of decision support based on estimation of group status by using conversation analysis},
 type = {book},
 year = {2015},
 source = {Communications in Computer and Information Science},
 keywords = {[Conversation estimation, Conversational agent, Gr},
 volume = {528},
 id = {1876071d-5f57-3248-859c-c28a72b538f5},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:46:58.389Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kono:HCII:2015},
 private_publication = {false},
 abstract = {© Springer International Publishing Switzerland 2015.We propose a model for a decision support of a group based on estimations of group status through utterance analysis. Based on methods used in prior studies of group dynamics and utterance analysis, we measured the utterance characteristics of group members to estimate group status; moreover, we aim to enhance the overall condition of the group by providing appropriate reference information in a timely manner through a conversational agent system. The goal of this model is a more satisfying decision-making process. Future work will focus on manufacturing a prototype system to verify both the operations involved in the test case and the ability to estimate group classification and status according to group dynamics.},
 bibtype = {book},
 author = {Kono, S. and Aihara, K.},
 doi = {10.1007/978-3-319-21380-4_106}
}

© Springer International Publishing Switzerland 2015.We propose a model for a decision support of a group based on estimations of group status through utterance analysis. Based on methods used in prior studies of group dynamics and utterance analysis, we measured the utterance characteristics of group members to estimate group status; moreover, we aim to enhance the overall condition of the group by providing appropriate reference information in a timely manner through a conversational agent system. The goal of this model is a more satisfying decision-making process. Future work will focus on manufacturing a prototype system to verify both the operations involved in the test case and the ability to estimate group classification and status according to group dynamics.

@book{
 title = {Spending precious travel time more wisely: A service model that provides instant travel assistance using input from locals},
 type = {book},
 year = {2015},
 source = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
 keywords = {[Behavior log, Context-aware computing, Cyberphysi},
 volume = {9189},
 id = {72bb75d4-63b5-33f0-ad90-846561a66757},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:41:21.126Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:HCII:2015},
 private_publication = {false},
 abstract = {© Springer International Publishing Switzerland 2015.We propose a new service model that incorporates input from locals to provide instant travel assistance. No matter how well travelers plan, circumstances at their destination invariably force them to modify their plans, especially towards the end of a tour. However, it is difficult for travelers to get useful information or support. Accordingly, we propose a service model whereby on-site travelers can instantly access planning suggestions from locals. Travelers send a personalized plan request with conditions such as desired end point (destination and time), preferences (e.g. dining, historical spots, etc.), and group information. The service delivers this to participating locals, who create appropriate plans and register them with the service. We conducted a preliminary survey of the model in March, 2014 in Matsuyama, Japan. Despite an insufficient volume of data the questionnaire results support our approach. Locals provided useful information that was perceived as more credible than online search results.},
 bibtype = {book},
 author = {Aihara, K. and Kono, S. and Sugino, S.},
 doi = {10.1007/978-3-319-20804-6_51}
}

© Springer International Publishing Switzerland 2015.We propose a new service model that incorporates input from locals to provide instant travel assistance. No matter how well travelers plan, circumstances at their destination invariably force them to modify their plans, especially towards the end of a tour. However, it is difficult for travelers to get useful information or support. Accordingly, we propose a service model whereby on-site travelers can instantly access planning suggestions from locals. Travelers send a personalized plan request with conditions such as desired end point (destination and time), preferences (e.g. dining, historical spots, etc.), and group information. The service delivers this to participating locals, who create appropriate plans and register them with the service. We conducted a preliminary survey of the model in March, 2014 in Matsuyama, Japan. Despite an insufficient volume of data the questionnaire results support our approach. Locals provided useful information that was perceived as more credible than online search results.

@inproceedings{
 title = {経路検索ログから抽出したイベントと交通状況との相関分析},
 type = {inproceedings},
 year = {2015},
 pages = {215-219},
 month = {12},
 publisher = {電子情報通信学会},
 id = {27c98c7e-0fab-32ee-ac0b-27eafcd5d3c8},
 created = {2025-08-02T16:39:40.749Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:39:52.672Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Piao:IEICE:2015},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {朴, 斌 and 相原, 健郎},
 booktitle = {信学技報}
}

@inproceedings{
 title = {会話分析によるグループ状態推定に基づく意思決定支援モデル},
 type = {inproceedings},
 year = {2015},
 pages = {101-106},
 month = {10},
 publisher = {電子情報通信学会},
 id = {deb4b3b4-2ad8-3142-a23d-ca0b855965e6},
 created = {2025-08-02T16:40:48.353Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:41:01.886Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kono:IEICE:2015},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {河野, 進 and 相原, 健郎},
 booktitle = {信学技報}
}

@inproceedings{
 title = {道路平面画像を用いた積雪環境下での路面幅推定},
 type = {inproceedings},
 year = {2015},
 month = {6},
 id = {818ee50c-99d1-3539-96a5-450d55745f4b},
 created = {2025-08-02T16:42:26.790Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:42:42.894Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Minematsu:SSII:2015},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {峰松, 翼 and 島田, 敬士 and 長原, 一 and 猪村, 元 and 相原, 健郎 and 谷口, 倫一郎},
 booktitle = {第21回画像センシングシンポジウム（SSII2015）}
}

@inproceedings{
 title = {省電力モバイルセンシングのための端末位置予測モデル },
 type = {inproceedings},
 year = {2015},
 month = {3},
 publisher = {電子情報通信学会},
 id = {1f55b082-29ea-36e4-9880-e333edc2f551},
 created = {2025-08-02T16:43:56.031Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:44:09.214Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kinoshita:IEICE:2015},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {木下, 僚 and 相原, 健郎 and 高須, 淳宏 and 石井, 淳 and 倉沢, 央 and 佐藤, 浩史 and 中村, 元紀 and 安達, 淳},
 booktitle = {2015年電子情報通信学会総合大会講演論文集}
}

@inproceedings{
 title = {街なかでの行動変容を促すサービス技術 〜センシングから情報提供まで〜},
 type = {inproceedings},
 year = {2015},
 month = {3},
 publisher = {電子情報通信学会},
 id = {51d23d37-462b-3fdd-9b47-0b34376a4bcb},
 created = {2025-08-02T16:45:02.986Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:45:12.939Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:IEICE:2015},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎},
 booktitle = {2015年電子情報通信学会総合大会講演論文集}
}

@inproceedings{
 title = {属性情報を利用した歩行者測位情報フィルタリング手法の提案},
 type = {inproceedings},
 year = {2015},
 pages = {181-186},
 publisher = {電子情報通信学会},
 id = {a5d012f7-dc31-3e30-baa6-26dd95485ca7},
 created = {2025-08-02T16:46:24.190Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:46:40.427Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Ishii:IEICE:2015},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {石井, 淳 and 倉沢, 央 and 佐藤, 浩史 and 中村, 元紀 and 高須, 淳宏 and 相原, 健郎 and 安達, 淳},
 booktitle = {信学技報}
}

@inproceedings{
 title = {位置情報サービスを通じた気分獲得を指向するモバイルセンシングの試み},
 type = {inproceedings},
 year = {2014},
 pages = {1-8},
 volume = {2014},
 issue = {5},
 id = {9f4cc89e-e26d-35ec-b68e-599754a85eb7},
 created = {2014-11-25T09:48:58.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:52:31.142Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:IPSJ:2014},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎},
 booktitle = {情報処理学会研究報告}
}

@article{
 title = {Collecting Behavior Logs with Emotions in Town},
 type = {article},
 year = {2014},
 keywords = {augmented environments,behavior log,context,emotion,human activity modelling and,sensor-,smart and hybrid cities,support},
 pages = {231-240},
 volume = {8530},
 websites = {http://link.springer.com/chapter/10.1007/978-3-319-07788-8_22},
 id = {99c1dd4a-33fe-3ef0-96b4-d0ca18874848},
 created = {2014-11-25T09:51:51.000Z},
 accessed = {2014-11-25},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:47:44.893Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {true},
 hidden = {false},
 citation_key = {Aihara:HCII:2014},
 private_publication = {false},
 abstract = {This paper proposes a new methodology for collecting visitors’ behavior and their emotions in town by using smartphones.
Existing social information services, such as Facebook and Twitter, are expanding to attach location data to users’ content. To capture situations of town, such as events what happens there or how people feel, the author believes that it’s not enough to collect tweets and behavior logs of locations in the town, because in fact the number of geotagged tweets is limited. Especially for microscopic analysis of town situations in small resolution of time and space, more information sources reflecting strollers’ behaviors and emotions are needed.
The paper proposed a function of LBS smartphone application to collect users’ behavior and emotions. When a user installs and uses an application with the function in town, the function records and transmits not only his/her locations but also his/her facial expressions by using front-facing camera.
An experiment was made in the beginning of November 2013. 55 subjects participated in the experiment. In addition to using the application in town, subjects were requested to provide correct data of facial expressions in 9 classes such as excited, fun and tired.
The function extracts 66 feature points of face by using Saragih’s model. As a quick result, the overall precision of 9 class-classification is 91.1% at 10-fold cross validation. The author believes that the result supports that the proposed application can collect facial expressions of not only active users who post microblogs but also read-only users.},
 bibtype = {article},
 author = {Aihara, Kenro},
 doi = {10.1007/978-3-319-07788-8_22},
 journal = {Distributed, Ambient, and Pervasive Interactions}
}

This paper proposes a new methodology for collecting visitors’ behavior and their emotions in town by using smartphones. Existing social information services, such as Facebook and Twitter, are expanding to attach location data to users’ content. To capture situations of town, such as events what happens there or how people feel, the author believes that it’s not enough to collect tweets and behavior logs of locations in the town, because in fact the number of geotagged tweets is limited. Especially for microscopic analysis of town situations in small resolution of time and space, more information sources reflecting strollers’ behaviors and emotions are needed. The paper proposed a function of LBS smartphone application to collect users’ behavior and emotions. When a user installs and uses an application with the function in town, the function records and transmits not only his/her locations but also his/her facial expressions by using front-facing camera. An experiment was made in the beginning of November 2013. 55 subjects participated in the experiment. In addition to using the application in town, subjects were requested to provide correct data of facial expressions in 9 classes such as excited, fun and tired. The function extracts 66 feature points of face by using Saragih’s model. As a quick result, the overall precision of 9 class-classification is 91.1% at 10-fold cross validation. The author believes that the result supports that the proposed application can collect facial expressions of not only active users who post microblogs but also read-only users.

@inproceedings{
 title = {Crowdsourced Mobile Sensing for Smarter City Life},
 type = {inproceedings},
 year = {2014},
 keywords = {Big data,Cities and towns,Crowdsourcing,Mobile communication,Roads,Sensors,Snow,Twitter,[behavior log, crowdsourcing, driving recorder, ex,behavior log,city administration,crowdsourced mobile sensing,crowdsourcing,driver information systems,driving recorder,exploratory visual analytics tool,interactive analysis,mobile computing,mobile sensing,mobile sensing framework,public administration,roadside phenomena,sensor data,service platform,smart phones,smarter city life,smartphone applications,social big data,social cyber-physical systems,social network services,social networking (online),video recording},
 pages = {334-337},
 publisher = {IEEE},
 id = {32be289f-644b-3e69-b5c2-00ac5b70fee8},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:53:01.238Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {true},
 hidden = {false},
 citation_key = {Aihara:SOCA:2014},
 private_publication = {false},
 abstract = {This paper introduces the ongoing project that aims to develop a mobile sensing framework to collect sensor data reflecting personal-scale, or microscopic, roadside phenomena by crowd sourcing and also using social big data, such as traffic, climate, and contents of social network services like Twitter. To collect them, smartphone applications are provided. One of the typical applications is a driving recorder that collects not only sensor data, such as acceleration, gyro, compass, and speed, but also recorded videos from the driver's view. To extract specific roadside phenomena, collected data are integrated and analyzed at the service platform. The framework also provides tools for interactive analysis to support city administration.},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Imura, Hajime and Takasu, Atsuhiro and Tanaka, Yuzuru and Adachi, Jun},
 doi = {10.1109/SOCA.2014.33},
 booktitle = {Proceedings of the 2014 IEEE 7th International Conference on Service-Oriented Computing and Applications Service-Oriented Computing and Applications}
}

This paper introduces the ongoing project that aims to develop a mobile sensing framework to collect sensor data reflecting personal-scale, or microscopic, roadside phenomena by crowd sourcing and also using social big data, such as traffic, climate, and contents of social network services like Twitter. To collect them, smartphone applications are provided. One of the typical applications is a driving recorder that collects not only sensor data, such as acceleration, gyro, compass, and speed, but also recorded videos from the driver's view. To extract specific roadside phenomena, collected data are integrated and analyzed at the service platform. The framework also provides tools for interactive analysis to support city administration.

@inproceedings{
 title = {訪問先での提示情報の受容性調査},
 type = {inproceedings},
 year = {2014},
 month = {9},
 id = {d6cbbb40-2c7f-3cd7-a9af-3a4b14e4ed93},
 created = {2025-08-02T16:47:44.295Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:48:03.415Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kono:FIT:2014},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {河野, 進 and 相原, 健郎},
 booktitle = {第13回情報科学技術フォーラム論文集}
}

@inproceedings{
 title = {The SMART-GS Project: An Approach to Image-based Digital Humanities},
 type = {inproceedings},
 year = {2014},
 websites = {https://dh-abstracts.library.virginia.edu/works/2112},
 month = {7},
 id = {97bd15ec-2bcd-361a-8fe0-8434e1a1ee61},
 created = {2025-08-02T16:49:32.566Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:48:48.768Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Hashimoto:DH:2014},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Hashimoto, Yuta and Aihara, Kenro and Hayashi, Susumu and Kukita, Minao and Ohura, Makoto},
 booktitle = {Proceedings of the Digital Humanities 2014}
}

@inproceedings{
 title = {行動ログとユーザインセンティブを考慮した地域LBSサービスモデルの一提案},
 type = {inproceedings},
 year = {2014},
 pages = {1-4},
 month = {5},
 publisher = {人工知能学会},
 id = {63191735-fd79-3c19-8980-8de6149e1a74},
 created = {2025-08-02T16:51:03.265Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:51:15.040Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:JSAI:2014},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎 and 杉野, 静弘},
 booktitle = {第28回 人工知能学会全国大会予稿集}
}

@inproceedings{
 title = {来訪者の反応を誘起するセンシングの提案},
 type = {inproceedings},
 year = {2014},
 pages = {1-3},
 month = {5},
 publisher = {人工知能学会},
 id = {489fd085-fb06-3afb-82bb-d13390d6cf51},
 created = {2025-08-02T16:51:58.618Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:52:12.901Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:JSAI:2014-2},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎},
 booktitle = {第28回 人工知能学会全国大会予稿集}
}

@article{
 title = { ユーザインセンティブと事業者メリットを考慮した旅行者向け位置情報サービス},
 type = {article},
 year = {2014},
 pages = {47-58},
 volume = {10},
 id = {4f33f5f1-d7c3-3c8e-b26b-4e096f77da25},
 created = {2025-08-02T16:54:32.944Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:54:56.950Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:STI:2014},
 private_publication = {false},
 bibtype = {article},
 author = {相原, 健郎 and 杉野, 静弘},
 journal = {観光情報学会誌「観光と情報」},
 number = {1}
}

@inproceedings{
 title = {Highway Incident Detection Based on Probe Car Data},
 type = {inproceedings},
 year = {2013},
 keywords = {automatic incident detection,big data,probe car data},
 pages = {103-110},
 id = {447fc1c8-b22e-3428-bd0a-c020aeccd28e},
 created = {2014-11-20T09:41:03.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {true},
 hidden = {false},
 private_publication = {false},
 abstract = {In this paper, we propose a method for automatic incident detection (AID) based on probe car data (PCD). Traffic incidents cause serious traffic congestion, which eventually leads to economic loss and environmental pollution. The early detection of a traffic incident is a very important component of traffic management. Many traffic incidents also occur during disasters, so the early detection of incidents is very important for disaster mitigation and damage limitation. From this perspective, PCD may be regarded as a cost-effective data source. A feature that can be obtained from PCD is different from that of road installation-type sensors. Thus, we investigated a feature that can be obtained from PCD to exploit PCD to the utmost. This feature is an index that represents the speed "deviation", which is extracted from each car's temporal and spatial mean speed and used for AID. We applied our method to data collected on days when incidents occurred. The evaluation results showed that our proposed method detected each incident correctly with an average false alarm rate of 0.08% or less. This shows that the method could be practiced effectively.},
 bibtype = {inproceedings},
 author = {Akatsuka, Hiroto and Takasu, Atsuhiro and Aihara, Kenro and Adachi, Jun},
 booktitle = {Proceedings of the International Conference on Information Systems 2013}
}

In this paper, we propose a method for automatic incident detection (AID) based on probe car data (PCD). Traffic incidents cause serious traffic congestion, which eventually leads to economic loss and environmental pollution. The early detection of a traffic incident is a very important component of traffic management. Many traffic incidents also occur during disasters, so the early detection of incidents is very important for disaster mitigation and damage limitation. From this perspective, PCD may be regarded as a cost-effective data source. A feature that can be obtained from PCD is different from that of road installation-type sensors. Thus, we investigated a feature that can be obtained from PCD to exploit PCD to the utmost. This feature is an index that represents the speed "deviation", which is extracted from each car's temporal and spatial mean speed and used for AID. We applied our method to data collected on days when incidents occurred. The evaluation results showed that our proposed method detected each incident correctly with an average false alarm rate of 0.08% or less. This shows that the method could be practiced effectively.

@inproceedings{
 title = {プローブカーデータを用いた高速道路における自動交通異常検出},
 type = {inproceedings},
 year = {2013},
 pages = {159-160},
 id = {38800a8c-9f86-3bfb-9321-737895b396e0},
 created = {2014-11-25T09:43:58.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:56:57.947Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Akatsuka:IPSJ:2013},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {赤塚, 裕人 and 高須, 淳宏 and 相原, 健郎 and 安達, 淳},
 booktitle = {情報処理学会第75回全国大会 予稿集}
}

@inproceedings{
 title = {Highway incident detection based on probe car data},
 type = {inproceedings},
 year = {2013},
 keywords = {[Automatic incident detection, Big data, Probe car},
 id = {897d35a1-dc09-301e-8ba9-f7b9890410f3},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:56:39.638Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Akatsuka:ICIS:2013},
 private_publication = {false},
 abstract = {© 2013 IADIS.In this paper, we propose a method for automatic incident detection (AID) based on probe car data (PCD). Traffic incidents cause serious traffic congestion, which eventually leads to economic loss and environmental pollution. The early detection of a traffic incident is a very important component of traffic management. Many traffic incidents also occur during disasters, so the early detection of incidents is very important for disaster mitigation and damage limitation. From this perspective, PCD may be regarded as a cost-effective data source. A feature that can be obtained from PCD is different from that of road installation-type sensors. Thus, we investigated a feature that can be obtained from PCD to exploit PCD to the utmost. This feature is an index that represents the speed "deviation", which is extracted from each car's temporal and spatial mean speed and used for AID. We applied our method to data collected on days when incidents occurred. The evaluation results showed that our proposed method detected each incident correctly with an average false alarm rate of 0.08% or less. This shows that the method could be practiced effectively.},
 bibtype = {inproceedings},
 author = {Akatsuka, H. and Takasu, A. and Aihara, K. and Adachi, J.},
 booktitle = {Proceedings of the IADIS International Conference Information Systems 2013, IS 2013}
}

© 2013 IADIS.In this paper, we propose a method for automatic incident detection (AID) based on probe car data (PCD). Traffic incidents cause serious traffic congestion, which eventually leads to economic loss and environmental pollution. The early detection of a traffic incident is a very important component of traffic management. Many traffic incidents also occur during disasters, so the early detection of incidents is very important for disaster mitigation and damage limitation. From this perspective, PCD may be regarded as a cost-effective data source. A feature that can be obtained from PCD is different from that of road installation-type sensors. Thus, we investigated a feature that can be obtained from PCD to exploit PCD to the utmost. This feature is an index that represents the speed "deviation", which is extracted from each car's temporal and spatial mean speed and used for AID. We applied our method to data collected on days when incidents occurred. The evaluation results showed that our proposed method detected each incident correctly with an average false alarm rate of 0.08% or less. This shows that the method could be practiced effectively.

@article{
 title = {Do Strollers in Town Needs Recommendation?: On Preferences of Recommender in Location-Based Services},
 type = {article},
 year = {2013},
 keywords = {behavioral cost,context-aware computing,location-based service},
 pages = {275-283},
 volume = {LNCS 8028},
 id = {bb7e3a3d-006e-3214-b8f0-b40a29d6ebcc},
 created = {2021-10-12T05:26:21.213Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T17:07:59.031Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:HCII:2013},
 private_publication = {false},
 abstract = {When we discuss about recommendation especially in Location-Based Services (LBS), we need to reveal whether users really want recommendations or not in fact while they are strolling in town, prior to evaluate each recommendation model. In this paper, a Location-Based Service, called nicotoco, is shown. nicotoco is an iPhone-based LBS in Futako-tamagawa area, Tokyo, Japan and provides information about stores and events to users. In the experiment using nicotoco, recommendations may be preferred more than rankings which was made from access counts.},
 bibtype = {article},
 author = {Aihara, Kenro},
 doi = {10.1007/978-3-642-39351-8_30},
 journal = {Distributed, Ambient, and Pervasive Interactions}
}

When we discuss about recommendation especially in Location-Based Services (LBS), we need to reveal whether users really want recommendations or not in fact while they are strolling in town, prior to evaluate each recommendation model. In this paper, a Location-Based Service, called nicotoco, is shown. nicotoco is an iPhone-based LBS in Futako-tamagawa area, Tokyo, Japan and provides information about stores and events to users. In the experiment using nicotoco, recommendations may be preferred more than rankings which was made from access counts.

@inproceedings{
 title = {混合ポアソンモデルによる道路状態推定},
 type = {inproceedings},
 year = {2013},
 pages = {679-680},
 id = {d740862f-fd52-3c40-a28a-22cddf802241},
 created = {2025-01-10T05:09:54.020Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-09-08T08:24:48.648Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Kinoshita:2013},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {木下, 僚 and 高須, 淳宏 and 安達, 淳},
 booktitle = {情報科学技術フォーラム講演論文集}
}

@inproceedings{
 title = {動体データによる旅行者の動向把握の可能性と課題},
 type = {inproceedings},
 year = {2013},
 pages = {17-20},
 month = {11},
 id = {b007cb2e-0c84-3826-b545-cd4200c3e018},
 created = {2025-08-02T16:55:53.319Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-02T16:56:06.714Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:STI:2013},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎 and 杉野, 静弘},
 booktitle = {観光情報学会 第8回研究発表会 講演論文集}
}

@inproceedings{
 title = {混合分布を用いた道路状態推定},
 type = {inproceedings},
 year = {2013},
 pages = {119-124},
 month = {1},
 publisher = {人工知能学会},
 id = {f4249936-13c1-3274-b567-c4658c36869b},
 created = {2025-08-03T11:59:11.228Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T11:59:45.118Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Yamamoto:JSAI:2013},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {山本, 敬介 and 安達, 淳 and 高須, 淳宏 and 相原, 健郎},
 booktitle = {人工知能学会第88回人工知能基本問題研究会予稿集}
}

@inproceedings{
 title = {街なかにおける気づきの設計について},
 type = {inproceedings},
 year = {2013},
 publisher = {人工知能学会},
 id = {fd991f51-602d-3d37-b6e1-054601716c61},
 created = {2025-08-03T12:01:23.433Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:01:57.687Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:JSAI:2013},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎 and 小柴, 等 and 杉野, 静弘 and 門倉, 博之},
 doi = {0.11517/pjsai.JSAI2013.0_1I4OS11b4},
 booktitle = {人工知能学会全国大会論文集}
}

@inproceedings{
 title = {街なかにおける目的外行動の誘起について},
 type = {inproceedings},
 year = {2012},
 publisher = {人工知能学会},
 id = {4a1b4a22-c64b-3a45-8e7a-039c77840e90},
 created = {2025-08-03T12:05:35.100Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:06:07.260Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Aihara:JSAI:2012},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎 and 小柴, 等 and 門倉, 博之 and 峰﨑, 大輔 and 金山, 明煥},
 doi = {10.11517/pjsai.JSAI2012.0_2P1OS9b2},
 booktitle = {人工知能学会全国大会論文集}
}

@inbook{
 type = {inbook},
 year = {2011},
 keywords = {[behavioral cost, context-aware computing, locatio},
 pages = {271-279},
 volume = {LNCS 6763},
 issue = {PART 3},
 websites = {http://link.springer.com/10.1007/978-3-642-21616-9_30},
 publisher = {Springer-Verlag},
 id = {d041a499-abbd-35b6-b26d-5a5f929ce4c3},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:17:13.251Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {This paper proposes a new model for recommendation based on the behavioral cost of recommendees in town. The model is based on cost-benefit analysis of the information provided to the user, referring to the model of temporal discounting and preference reversal. Here we assume that behavioral cost may be regarded as time in temporal discounting. A recommender system based on this model can select information, which is located in the surrounding area (not so far away) and may be preferred by the user, if the system can estimate where the reversal phenomenon may occur. The experiments were made using an experimental social service, called "pin@clip", which is an iPhone-based social bookmarking service in Shibuya, Tokyo, Japan that has been operating since December 2009. The experimental results show that the phenomenon of preference reversals might occur, even though the authors could not obtain statistically significant data. © 2011 Springer-Verlag.},
 bibtype = {inbook},
 author = {Aihara, Kenro and Koshiba, Hitoshi and Takeda, Hideaki},
 doi = {10.1007/978-3-642-21616-9_30},
 chapter = {Behavioral Cost-Based Recommendation Model for Wanderers in Town},
 title = {Human-Computer Interaction, Part III, HCII 2011}
}

This paper proposes a new model for recommendation based on the behavioral cost of recommendees in town. The model is based on cost-benefit analysis of the information provided to the user, referring to the model of temporal discounting and preference reversal. Here we assume that behavioral cost may be regarded as time in temporal discounting. A recommender system based on this model can select information, which is located in the surrounding area (not so far away) and may be preferred by the user, if the system can estimate where the reversal phenomenon may occur. The experiments were made using an experimental social service, called "pin@clip", which is an iPhone-based social bookmarking service in Shibuya, Tokyo, Japan that has been operating since December 2009. The experimental results show that the phenomenon of preference reversals might occur, even though the authors could not obtain statistically significant data. © 2011 Springer-Verlag.

@inproceedings{
 title = {街なかでのユーザ体験向上とログ収集を目的とするLocation-Based Service − pin@clip（ピクナリ）サービスとその基盤 −},
 type = {inproceedings},
 year = {2011},
 month = {11},
 id = {ebf82484-a15d-305b-bb25-6524e308ef1b},
 created = {2025-08-03T12:07:48.884Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:07:58.396Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎 and 小柴, 等 and 門倉, 博之 and 峰崎, 大輔 and 金山, 明煥},
 booktitle = {東京大学 空間情報科学研究センター シンポジウム CSIS DAYS 2011 研究アブストラクト集}
}

@inproceedings{
 title = {ARマーカと行動ログを活用した地域活性化プラットフォームの考察 : e空間実現のためのサービス実証実験pin@clip 2010},
 type = {inproceedings},
 year = {2011},
 pages = {615-620},
 month = {9},
 publisher = {FIT（電子情報通信学会・情報処理学会)運営委員会）},
 day = {7},
 id = {e0c35520-eb9b-3c3a-808b-a7faa55be77f},
 created = {2025-08-03T12:10:06.952Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:10:15.637Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {小柴, 等 and 相原, 健郎 and 門倉, 博之 and 峰崎, 大輔 and 金山, 明煥},
 booktitle = {情報科学技術フォーラム講演論文集}
}

@inproceedings{
 title = {画像化主義に基づく文献資料研究用ツールSMART-GSとその発展},
 type = {inproceedings},
 year = {2011},
 pages = {1-5},
 month = {1},
 publisher = {情報処理学会},
 id = {4653ecd2-733b-38cf-b7c0-9d5d81e05c76},
 created = {2025-08-03T12:14:48.018Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:14:52.844Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {相原, 健郎 and 林, 晋},
 booktitle = {情報処理学会 デジタル・ドキュメント研究会研究報告}
}

@inproceedings{
 title = {街なかソーシャル・ブックマーキング "pin@clip ピナクリ" — e空間実現のためのサービス実証実験の全体像 —},
 type = {inproceedings},
 year = {2010},
 pages = {417-420},
 month = {9},
 id = {d73f9db8-ae5e-393b-8110-3fd75e62920a},
 created = {2025-08-03T12:22:07.816Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:22:30.022Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {中尾, 敏康 and 相原, 健郎 and 小方, 靖 and 田代, 光平 and 小柴, 等 and 宮崎, 陽司 and 小西, 勇介 and 武田, 英明 and 佐々木, 憲二 and 金山, 明煥},
 booktitle = {第9回情報科学技術フォーラム 講演論文集}
}

@article{
 title = {説得性に基づく情報推薦手法の提案：送り手の属性に着目したモデルと検証},
 type = {article},
 year = {2010},
 pages = {1452-1468},
 volume = {51},
 month = {8},
 id = {f1e4cd58-2dff-35e8-a6e5-6a2e2e9d56da},
 created = {2025-08-03T12:45:04.439Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:51:36.463Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Koshiba:JSAI:2010},
 private_publication = {false},
 abstract = {ICT（Information and Communication Technology）の発展・普及にともなって，情報爆発という問題が顕在化してきた1）．この情報爆発と呼ばれる問題への対策の1つとして，ユーザの望む，ユーザに適した情報を提供するための情報推薦技術が研究・開発されている．しかしながら，情報爆発の勢いはとどまることを知らず，すでに推薦結果であるユーザに適した情報までが氾濫しつつある．そこで本論文では，ネット上や物理世界における行動ログデータが多数蓄積された世界を想定し，従来の推薦手法に加えて，推薦結果である情報がユーザに受け入れられる可能性（情報の受容度）まで考慮した情報推薦手法を提案し，実験により評価した．具体的には，まず，社会心理学の知見をベースに説得性と称するユーザ間の情報受容度を算出するモデルを提案した．次に，実際の商店街を舞台に一般募集の被験者を対象として，行動ログデータを収集した．そのうえで，それらの行動ログデータを用いて説得性の算出を行い，被験者による評価実験を行った．評価実験の結果は，おおむね我々の仮説を好意的に支持するものであり，本論文において提案した“説得性を考慮した情報推薦手法”が，有効に作用することが確認できた．

In the “information explosion” are when we can access the increasing amount of information available in digital form from anywhere and at anytime, a filtering method is one of the most important technologies to be solved; how to identify information whether it is adequate and relevant to a user or not. Although many information recommendation systems have been proposed, existing methods could not solve the problem of “overflow” of recommendations, if they judge the information only by using similarity measures of contents. In this paper, we propose a new recommendation approach based on persuasibility among users. We focus on lifelogging both in the real world and the information world. Our recommendation model the information acceptability of information from unknown users and assume that the information of highly persuasible users tend to be accepted than ones of little persuasible users. Our persuasibility model consists of expertise, trustworthiness, likability, and similarity which is typically used in collaborative filtering methods. We carried out some experiments for evaluating our method. Firstly, we developed a specific recommendation model and an experimental recommender. Secondly, we set an experiment to collect lifelogs in a authentic situation. In this experiment, we used 52 shops in a small but very popular town in Tokyo, and about 700 consumers were involved. Finally, we set up another experiment for evaluation of our recommender using a subset of these participants. Our experimental results suggest that our proposed “persuasibility-based recommender” is more useful than general collaborative filtering.},
 bibtype = {article},
 author = {小柴, 等 and 相原, 健郎 and 小田, 朋宏 and 星, 孝哲 and 松原, 伸人 and 森, 純一郎 and 武田, 英明},
 journal = {情報処理学会論文誌},
 number = {8}
}

ICT（Information and Communication Technology）の発展・普及にともなって，情報爆発という問題が顕在化してきた1）．この情報爆発と呼ばれる問題への対策の1つとして，ユーザの望む，ユーザに適した情報を提供するための情報推薦技術が研究・開発されている．しかしながら，情報爆発の勢いはとどまることを知らず，すでに推薦結果であるユーザに適した情報までが氾濫しつつある．そこで本論文では，ネット上や物理世界における行動ログデータが多数蓄積された世界を想定し，従来の推薦手法に加えて，推薦結果である情報がユーザに受け入れられる可能性（情報の受容度）まで考慮した情報推薦手法を提案し，実験により評価した．具体的には，まず，社会心理学の知見をベースに説得性と称するユーザ間の情報受容度を算出するモデルを提案した．次に，実際の商店街を舞台に一般募集の被験者を対象として，行動ログデータを収集した．そのうえで，それらの行動ログデータを用いて説得性の算出を行い，被験者による評価実験を行った．評価実験の結果は，おおむね我々の仮説を好意的に支持するものであり，本論文において提案した“説得性を考慮した情報推薦手法”が，有効に作用することが確認できた． In the “information explosion” are when we can access the increasing amount of information available in digital form from anywhere and at anytime, a filtering method is one of the most important technologies to be solved; how to identify information whether it is adequate and relevant to a user or not. Although many information recommendation systems have been proposed, existing methods could not solve the problem of “overflow” of recommendations, if they judge the information only by using similarity measures of contents. In this paper, we propose a new recommendation approach based on persuasibility among users. We focus on lifelogging both in the real world and the information world. Our recommendation model the information acceptability of information from unknown users and assume that the information of highly persuasible users tend to be accepted than ones of little persuasible users. Our persuasibility model consists of expertise, trustworthiness, likability, and similarity which is typically used in collaborative filtering methods. We carried out some experiments for evaluating our method. Firstly, we developed a specific recommendation model and an experimental recommender. Secondly, we set an experiment to collect lifelogs in a authentic situation. In this experiment, we used 52 shops in a small but very popular town in Tokyo, and about 700 consumers were involved. Finally, we set up another experiment for evaluation of our recommender using a subset of these participants. Our experimental results suggest that our proposed “persuasibility-based recommender” is more useful than general collaborative filtering.

@article{
 title = {記憶の想起と記録のためのライフログ・ブログ連携型支援手法の提案},
 type = {article},
 year = {2010},
 pages = {63-81},
 volume = {51},
 id = {73162a73-85ea-3de5-af7e-46634ac158e6},
 created = {2025-08-03T12:48:14.697Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:52:37.891Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {本論文では, ブログとライフログを利用した記憶の想起と記録の支援システムの提案と, 当該システムの一般ユーザにおける受容性について述べる. まず, ブログやライフログを巡る現状の課題に関する概観について述べ, それらの課題を解決するために, ライフログデータをもとにしたブログ生成支援システムの提案と, 支援システムを支える, ライフログデータからのユーザの興味・関心推定手法の提案を行う. 次いで, 実店舗と一般ユーザを用いた実証実験により, 当該システムの有効性と受容性について考察を行う.
実証実験の結果からは, ライフログデータをもとにしたブログ生成支援システムについて, 普段ブログに関する活動に積極的でないユーザからも多くの記事が編集されるなど, 広く受容され, 有効に機能したこと, 筆者らの提案する興味・関心推定手法は, 単純に滞在時間によって興味・関心を推定する手法に比べて, より高精度に興味・関心を推定できること, ライフログデータをもとにブログ記事のひな形を作成する際に, ユーザの興味・関心という主観を用いてデータを丸めることで, 行動情報の取得に対する嫌悪感が低減され受容されやすくなる可能性のあること, などを示唆するデータを得た.},
 bibtype = {article},
 author = {小柴, 等 and 相原, 健郎 and 森, 純一郎 and 小田, 朋宏 and 星, 孝哲 and 松原, 伸人 and 武田, 英明},
 journal = {情報処理学会論文誌},
 number = {1}
}

本論文では, ブログとライフログを利用した記憶の想起と記録の支援システムの提案と, 当該システムの一般ユーザにおける受容性について述べる. まず, ブログやライフログを巡る現状の課題に関する概観について述べ, それらの課題を解決するために, ライフログデータをもとにしたブログ生成支援システムの提案と, 支援システムを支える, ライフログデータからのユーザの興味・関心推定手法の提案を行う. 次いで, 実店舗と一般ユーザを用いた実証実験により, 当該システムの有効性と受容性について考察を行う. 実証実験の結果からは, ライフログデータをもとにしたブログ生成支援システムについて, 普段ブログに関する活動に積極的でないユーザからも多くの記事が編集されるなど, 広く受容され, 有効に機能したこと, 筆者らの提案する興味・関心推定手法は, 単純に滞在時間によって興味・関心を推定する手法に比べて, より高精度に興味・関心を推定できること, ライフログデータをもとにブログ記事のひな形を作成する際に, ユーザの興味・関心という主観を用いてデータを丸めることで, 行動情報の取得に対する嫌悪感が低減され受容されやすくなる可能性のあること, などを示唆するデータを得た.

@inbook{
 type = {inbook},
 year = {2009},
 pages = {126-137},
 volume = {LNCS 5535},
 websites = {http://link.springer.com/10.1007/978-3-642-02247-0_14},
 publisher = {Springer Berlin Heidelberg},
 id = {78e960a8-023f-3f16-a287-e24ef8bf0eb5},
 created = {2014-11-20T09:41:03.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:55:50.980Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {true},
 hidden = {false},
 citation_key = {Mori2009},
 private_publication = {false},
 abstract = {Recent sensor technologies have enabled the capture of users’ behavior data. Given the large amount of data currently available from sensor-equipped environments, it is important to attempt characterization of the sensor data for automatically modeling users in a ubiquitous and mobile computing environment. As described herein, we propose a method that predicts a customer model using features based on customers’ behavior in a shop. We capture the customers’ behavior using various sensors in the form of the time duration and the sequence between blocks in the shop. Based on behavior data from the sensors, we design features that characterize the behavior pattern of a customer in the shop. We employ those features using a machine learning approach to predict customer attributes such as age, gender, occupation, and interest. Our results show that our designed behavior-based features perform with F-values of 70–90% for prediction. We also discuss the potential applications of our method in user modeling.},
 bibtype = {inbook},
 author = {Mori, Junichiro and Matsuo, Yutaka and Koshiba, Hitoshi and Aihara, Kenro and Takeda, Hideaki},
 doi = {10.1007/978-3-642-02247-0_14},
 chapter = {Predicting Customer Models Using Behavior-Based Features in Shops},
 title = {User Modeling, Adaptation, and Personalization}
}

Recent sensor technologies have enabled the capture of users’ behavior data. Given the large amount of data currently available from sensor-equipped environments, it is important to attempt characterization of the sensor data for automatically modeling users in a ubiquitous and mobile computing environment. As described herein, we propose a method that predicts a customer model using features based on customers’ behavior in a shop. We capture the customers’ behavior using various sensors in the form of the time duration and the sequence between blocks in the shop. Based on behavior data from the sensors, we design features that characterize the behavior pattern of a customer in the shop. We employ those features using a machine learning approach to predict customer attributes such as age, gender, occupation, and interest. Our results show that our designed behavior-based features perform with F-values of 70–90% for prediction. We also discuss the potential applications of our method in user modeling.

@inbook{
 type = {inbook},
 year = {2009},
 keywords = {[Interface design, Knowledge acquisition, Motion c},
 pages = {51-59},
 volume = {LNCS 5622},
 websites = {http://link.springer.com/10.1007/978-3-642-02771-0_6},
 publisher = {Springer-Verlag},
 id = {7382347c-ff42-31b3-b5ac-cd43a796b4e0},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:53:34.257Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 citation_key = {Fujisawa:HCII:2009},
 private_publication = {false},
 abstract = {We propose a methodology for estimating a user's interest in documents displayed on a computer screen from his or her physical actions. Some studies show that physical actions captured by a device can be indicators of a user's interest. We introduce the ongoing pilot study's results, which show the possible relationship between a user's face approaching the screen, as captured by a webcam, and their interest in the document on the screen. Our system uses a common user-friendly device. We evaluate our prototype system from the viewpoint of presuming an interest from such a face approach and the practicality of the system, and discuss the future possibilities of our research. © 2009 Springer Berlin Heidelberg.},
 bibtype = {inbook},
 author = {Fujisawa, Kumiko and Aihara, Kenro},
 doi = {10.1007/978-3-642-02771-0_6},
 chapter = {Estimation of User Interest from Face Approaches Captured by Webcam},
 title = {Virtual and Mixed Reality}
}

We propose a methodology for estimating a user's interest in documents displayed on a computer screen from his or her physical actions. Some studies show that physical actions captured by a device can be indicators of a user's interest. We introduce the ongoing pilot study's results, which show the possible relationship between a user's face approaching the screen, as captured by a webcam, and their interest in the document on the screen. Our system uses a common user-friendly device. We evaluate our prototype system from the viewpoint of presuming an interest from such a face approach and the practicality of the system, and discuss the future possibilities of our research. © 2009 Springer Berlin Heidelberg.

@inproceedings{
 title = {言い換え箇所と言い換え候補の提示による解説文リライト支援 -読み手の評価実験},
 type = {inproceedings},
 year = {2009},
 publisher = {情報処理学会},
 id = {73ae4442-afd2-3d09-8681-a645c55fc5f9},
 created = {2025-08-03T12:50:38.984Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:50:43.391Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {藤沢, 仁子 and 神門, 典子 and 相原, 健郎 and 安達, 淳},
 booktitle = {情報処理学会人文科学とコンピュータシンポジウム論文集}
}

@inproceedings{
 title = {説得性に基づく情報推薦手法の提案 − 「ぷらっとPlat@自由が丘」における統合された行動ログの活用 −},
 type = {inproceedings},
 year = {2009},
 pages = {31-42},
 month = {9},
 id = {ef5bfe67-04b8-33ee-be37-5f3a1895c545},
 created = {2025-08-03T12:52:40.473Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T12:52:44.749Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {小柴, 等 and 相原, 健郎 and 森, 純一郎 and 武田, 英明 and 小田, 朋宏 and 星, 孝哲 and 松原, 伸人},
 booktitle = {第8回情報科学技術フォーラム 講演論文集}
}

@inproceedings{
 title = {Translation of on-screen text into visual expressions},
 type = {inproceedings},
 year = {2008},
 keywords = {[Data management, HCI, Visualization]},
 pages = {451-451},
 websites = {https://dl.acm.org/doi/10.1145/1378889.1378996},
 month = {6},
 publisher = {ACM},
 day = {16},
 city = {New York, NY, USA},
 id = {2368b1a9-c8e5-3530-8ce0-f64985b39a52},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:02:31.715Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {To support the user's on-screen reading, we propose a methodology to translate text into visual expressions. Our prototype system analyses the text and translates them into image and adds movement to the image.},
 bibtype = {inproceedings},
 author = {Fujisawa, Kumiko and Aihara, Kenro},
 doi = {10.1145/1378889.1378996},
 booktitle = {Proceedings of the 8th ACM/IEEE-CS joint conference on Digital libraries}
}

To support the user's on-screen reading, we propose a methodology to translate text into visual expressions. Our prototype system analyses the text and translates them into image and adds movement to the image.

@inproceedings{
 title = {Information extraction from scanned documents by stochastic page layout analysis},
 type = {inproceedings},
 year = {2008},
 pages = {447-448},
 websites = {https://dl.acm.org/doi/10.1145/1363686.1363795},
 month = {3},
 publisher = {ACM},
 day = {16},
 city = {New York, NY, USA},
 id = {e97d6d11-5d6e-3d76-9401-5dfc43aaab6b},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:02:49.975Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {We propose a stochastic context-free grammar for extracting information from scanned document images. The grammar is designed to disambiguate layout analysis and utilize both layout and text features. We applied this grammar to the problem of extracting bibliographic information from scanned academic papers and found that it can accurately extract information. Copyright 2008 ACM.},
 bibtype = {inproceedings},
 author = {Takasu, Atsuhiro and Aihara, Kenro},
 doi = {10.1145/1363686.1363795},
 booktitle = {Proceedings of the 2008 ACM symposium on Applied computing}
}

We propose a stochastic context-free grammar for extracting information from scanned document images. The grammar is designed to disambiguate layout analysis and utilize both layout and text features. We applied this grammar to the problem of extracting bibliographic information from scanned academic papers and found that it can accurately extract information. Copyright 2008 ACM.

@inproceedings{
 title = {Supporting Creation and Sharing of Contents of Cultural Heritage Objects for Educational Purposes},
 type = {inproceedings},
 year = {2008},
 pages = {5-14},
 month = {7},
 id = {32c452eb-6b51-3334-9ae9-6cf20af4836d},
 created = {2025-08-03T13:01:53.796Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:02:09.532Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Yamada, Taizo and Kando, Noriko and Fujisawa, Satoko and Uehara, Yusuke and Baba, Takayuki and Nagata, Shigemi and Tojo, Takashi and Adachi, Jun},
 booktitle = {Proceedings of the workshop on Personalized Access to Cultural Heritage (PATCH 2008)}
}

@inproceedings{
 title = {A smoothing method for a statistical string similarity},
 type = {inproceedings},
 year = {2007},
 pages = {624-629},
 id = {bf86e7ea-22f0-382c-94d0-cba406c1fb99},
 created = {2014-11-20T09:41:03.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {true},
 hidden = {false},
 private_publication = {false},
 abstract = {We often need to measure similarity between objective information when integrating information. We propose an algorithm in this paper for the Bayesian estimation of the parameters of a statistical string similarity model. To do this we need a smoothing technique for the parameter estimation, because a string similarity model usually contains many parameters. We can make a parameter estimation for the statistical similarity model by introducing a Dirichlet prior. The experimental results show that the proposed method is effective enough for approximate matching.},
 bibtype = {inproceedings},
 author = {Takasu, Atsuhiro and Aihara, Kenro and Yamada, Taizo},
 booktitle = {2007 IEEE International Conference on Information Reuse and Integration, IEEE IRI-2007}
}

We often need to measure similarity between objective information when integrating information. We propose an algorithm in this paper for the Bayesian estimation of the parameters of a statistical string similarity model. To do this we need a smoothing technique for the parameter estimation, because a string similarity model usually contains many parameters. We can make a parameter estimation for the statistical similarity model by introducing a Dirichlet prior. The experimental results show that the proposed method is effective enough for approximate matching.

@inproceedings{
 title = {A methodology of translation of sentences into iconic expressions},
 type = {inproceedings},
 year = {2007},
 keywords = {[And natural language processing, Human-computer i},
 id = {5bb69e2c-120f-39b7-b414-bf48bd654ea0},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {We propose a tool that converts sentences into visualized expressions of a subject and its action that supports the comprehension and memorization of their content. This system was applied to sentences excerpted from an actual story, and its visualization function was evaluated from the viewpoints of sentence and iconic expressions, the aggregation of verbs, and image retrieval.},
 bibtype = {inproceedings},
 author = {Fujisawa, K. and Aihara, K.},
 booktitle = {Proceedings of the 11th IASTED International Conference on Internet and Multimedia Systems and Applications, IMSA 2007}
}

We propose a tool that converts sentences into visualized expressions of a subject and its action that supports the comprehension and memorization of their content. This system was applied to sentences excerpted from an actual story, and its visualization function was evaluated from the viewpoints of sentence and iconic expressions, the aggregation of verbs, and image retrieval.

@inproceedings{
 title = {Building a Space of Cultural Heritage Objects to Explore It in the Classroom},
 type = {inproceedings},
 year = {2007},
 id = {fcf16dac-bdc5-3eec-9d54-8753fd82c09b},
 created = {2025-08-03T13:05:26.785Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:54:44.538Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Not only for further understanding of subjects but also for learning analytical thinking or inductive inference, it is important to provide a highly engageable environment where pupils can interact with high quality cultural contents easily. The authors have been developed an interactive exploring tool called CEAX Voyager. One of the important issues to make the learning with CEAX Voyager effective is to prepare adaptive descriptions for pupils. This paper proposes a framework for supporting creation and sharing of adaptive descriptions about cultural heritage objects. In the framework, such descriptions can be authored not only by professionals such as curators but also by school teachers.},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Yamada, Taizo and Kando, Noriko and Fujisawa, Satoko and Uehara, Yusuke and Baba, Takayuki and Nagata, Shigemi and Tojo, Takashi and Awaji, Tetsuhiko and Adachi, Jun},
 booktitle = {Proceedings of the 10th International Conference on Interactive Computer-Aided Learning (ICL2007)}
}

Not only for further understanding of subjects but also for learning analytical thinking or inductive inference, it is important to provide a highly engageable environment where pupils can interact with high quality cultural contents easily. The authors have been developed an interactive exploring tool called CEAX Voyager. One of the important issues to make the learning with CEAX Voyager effective is to prepare adaptive descriptions for pupils. This paper proposes a framework for supporting creation and sharing of adaptive descriptions about cultural heritage objects. In the framework, such descriptions can be authored not only by professionals such as curators but also by school teachers.

@inproceedings{
 title = {An annotation method for sensor data streams based on statistical patterns},
 type = {inproceedings},
 year = {2006},
 keywords = {[Engineering database, Hidden Markov model, Inform},
 id = {5cf24cf6-a8ea-3bbf-8f1f-2f43e19ec46d},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Due to recent advances in sensor technology greater quantities of sensor data are being generated and circulated. In these circumstances, sensor data stream processing and management technologies have become important research areas. In the development of mechanical and electrical systems, sensor stream data are a potential medium for sharing information among the engineers who are engaged in the various phases in the system development and operation. This paper proposes an annotation method for a sensor data stream that links the information generated in the development and operational phases of a system. The key techniques of the proposed method are sensor pattern construction using hidden Markov models (HMMs) and an annotation method based on the HMMs constructed. We applied the proposed method to the sensor data stream of a supersmall artificial satellite and showed that the proposed method achieved approximately 95% annotation accuracy for long fragments of the sensor data stream.},
 bibtype = {inproceedings},
 author = {Takasu, A. and Aihara, K.},
 booktitle = {Proceedings of the IASTED International Conference on Databases and Applications, DBA 2006}
}

Due to recent advances in sensor technology greater quantities of sensor data are being generated and circulated. In these circumstances, sensor data stream processing and management technologies have become important research areas. In the development of mechanical and electrical systems, sensor stream data are a potential medium for sharing information among the engineers who are engaged in the various phases in the system development and operation. This paper proposes an annotation method for a sensor data stream that links the information generated in the development and operational phases of a system. The key techniques of the proposed method are sensor pattern construction using hidden Markov models (HMMs) and an annotation method based on the HMMs constructed. We applied the proposed method to the sensor data stream of a supersmall artificial satellite and showed that the proposed method achieved approximately 95% annotation accuracy for long fragments of the sensor data stream.

@inproceedings{
 title = {A platform for managing term dictionaries for utilizing distributed interview archives},
 type = {inproceedings},
 year = {2006},
 volume = {239},
 id = {c0cf0af0-cbee-3d16-8795-dd71123e92dc},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {This paper proposes a platform that aims to support the whole process and facilitate archiving tasks at museums and galleries. When we try to preserve tacit knowledge or skills of artists or masters by interview and utilize interview videos, it is important to support the whole process of archiving; capturing interview, digitizing it to a machine readable format, authorizing metadata, and exporting them to be accessed from others. To interoperate distributed archives, rich meta information is required. In this paper, we focus our tools to facilitate authoring metadata of interview, such as transcription and annotation task for interview, with speech processing or natural language processing functions. The important point now is that the performance of the speech processing or text processing for specific domain depends on well-fitted language models to it. We, therefore, design our tools as an networked software package that can download up-to-date term dictionary and language models for speech processing or text processing from a server of the platform. Users can utilize tools with updated models. Meanwhile, created metadata including terms occurred in interview can be uploaded to the server and they will be used for update of the models hereafter. In this paper, we first overview our project. Then, we propose an platform and our tools to be distributed.},
 bibtype = {inproceedings},
 author = {Aihara, K. and Takasu, A.},
 booktitle = {CEUR Workshop Proceedings}
}

This paper proposes a platform that aims to support the whole process and facilitate archiving tasks at museums and galleries. When we try to preserve tacit knowledge or skills of artists or masters by interview and utilize interview videos, it is important to support the whole process of archiving; capturing interview, digitizing it to a machine readable format, authorizing metadata, and exporting them to be accessed from others. To interoperate distributed archives, rich meta information is required. In this paper, we focus our tools to facilitate authoring metadata of interview, such as transcription and annotation task for interview, with speech processing or natural language processing functions. The important point now is that the performance of the speech processing or text processing for specific domain depends on well-fitted language models to it. We, therefore, design our tools as an networked software package that can download up-to-date term dictionary and language models for speech processing or text processing from a server of the platform. Users can utilize tools with updated models. Meanwhile, created metadata including terms occurred in interview can be uploaded to the server and they will be used for update of the models hereafter. In this paper, we first overview our project. Then, we propose an platform and our tools to be distributed.

@inproceedings{
 title = {Quality enhancement in information extraction from scanned documents},
 type = {inproceedings},
 year = {2006},
 keywords = {[Digital libraries, Document capturing, Statistica},
 volume = {2006},
 id = {0bfaac2c-edb9-3e36-a1da-be9e49cd9036},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {When constructing a large document archive, an important element is the digitizing of printed documents. Although various techniques for document image analysis such as Optical Character Recognition (OCR) have been developed, error handling is required in constructing real document archive systems. This paper discusses the problem from the quality enhancement perspective and proposes a robust reference extraction method for academic articles scanned with OCR mark-up. We applied the proposed method to articles appearing in various journals, and these experiments showed that the proposed method achieved a recognition accuracy of more than 94%. This paper also discusses manual correction and investigates experimentally the relationship between extraction accuracy and cost reduction. Copyright 2006 ACM.},
 bibtype = {inproceedings},
 author = {Takasu, A. and Aihara, K.},
 doi = {10.1145/1166160.1166193},
 booktitle = {Proceedings of the 2006 ACM Symposium on Document Engineering, DocEng 2006}
}

When constructing a large document archive, an important element is the digitizing of printed documents. Although various techniques for document image analysis such as Optical Character Recognition (OCR) have been developed, error handling is required in constructing real document archive systems. This paper discusses the problem from the quality enhancement perspective and proposes a robust reference extraction method for academic articles scanned with OCR mark-up. We applied the proposed method to articles appearing in various journals, and these experiments showed that the proposed method achieved a recognition accuracy of more than 94%. This paper also discusses manual correction and investigates experimentally the relationship between extraction accuracy and cost reduction. Copyright 2006 ACM.

@inbook{
 type = {inbook},
 year = {2006},
 pages = {22-31},
 volume = {LNCS 4312},
 websites = {http://link.springer.com/10.1007/11931584_5},
 publisher = {Springer},
 city = {Berlin, Heidelberg},
 id = {55ad6279-8376-3c55-993e-21f832a43ee5},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:57:44.093Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {With the Educational use of Cultural heritage Archives and Cross(X) search (CEAX), we have investigated how to establish a framework for managing various kinds of information on cultural heritage objects and how to utilize them for educational purposes. To achieve this goal, we propose a conceptual framework in this paper called "Growing Metadata" and a flexible content management system called Owlery. Growing Metadata includes not only factual descriptions of objects but also various annotations about the objects, such as metadata for children, course materials prepared by school teachers, classroom reports, etc., and are reusable for search and educational purposes. Owlery is a software platform to create, share, utilize and reuse the Growing Metadata, and in which various metadata and annotations are managed in different levels of authenticity, authorship, and user groups. As a result of the experimental classes for 89 6th-grade children, our framework was found to be efficient and accepted by the content creators, like museum experts, content annotators and shool teachers. © Springer-Verlag Berlin Heidelberg 2006.},
 bibtype = {inbook},
 author = {Aihara, Kenro and Yamada, Taizo and Kando, Noriko and Fujisawa, Satoko and Uehara, Yusuke and Baba, Takayuki and Nagata, Shigemi and Tojo, Takashi and Awaji, Tetsuhiko and Adachi, Jun},
 doi = {10.1007/11931584_5},
 chapter = {Owlery: A Flexible Content Management System for “Growing Metadata” of Cultural Heritage Objects and Its Educational Use in the CEAX Project},
 title = {Digital Libraries: Achievements, Challenges and Opportunities. ICADL 2006}
}

With the Educational use of Cultural heritage Archives and Cross(X) search (CEAX), we have investigated how to establish a framework for managing various kinds of information on cultural heritage objects and how to utilize them for educational purposes. To achieve this goal, we propose a conceptual framework in this paper called "Growing Metadata" and a flexible content management system called Owlery. Growing Metadata includes not only factual descriptions of objects but also various annotations about the objects, such as metadata for children, course materials prepared by school teachers, classroom reports, etc., and are reusable for search and educational purposes. Owlery is a software platform to create, share, utilize and reuse the Growing Metadata, and in which various metadata and annotations are managed in different levels of authenticity, authorship, and user groups. As a result of the experimental classes for 89 6th-grade children, our framework was found to be efficient and accepted by the content creators, like museum experts, content annotators and shool teachers. © Springer-Verlag Berlin Heidelberg 2006.

@inproceedings{
 title = {CEAX's Learning Support System to Explore Cultural Heritage Objects without Keyword Search},
 type = {inproceedings},
 year = {2006},
 month = {11},
 city = {Seville, Spain},
 id = {36c2ca90-3f45-3438-ad97-1a786946e97d},
 created = {2025-08-03T13:08:59.183Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:09:06.817Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Yamada, Taizo and Aihara, Kenro and Kando, Noriko and Fujisawa, Satoko and Uehara, Yusuke and Baba, Takayuki and Nagata, Shigemi and Tojo, Takashi and Hiroshima, Yuko and Adachi, Jun},
 booktitle = {the fourth International Conference on Multimedia and Information and Communication Technologies in Education}
}

@inproceedings{
 title = {A reciprocal platform for archiving interview videos about arts and crafts},
 type = {inproceedings},
 year = {2005},
 keywords = {[Interview video, Knowledge sharing, NLP resource},
 id = {fcf14ec0-5f71-35dc-853a-0098577764ce},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {This paper proposes a platform for portal and local repositories. Our methodology aims not only at construction of portal site but also at supporting capture of digital contents transformed from interview videos with intellectuals.},
 bibtype = {inproceedings},
 author = {Aihara, K. and Takasu, A.},
 booktitle = {Proceedings of the ACM/IEEE Joint Conference on Digital Libraries}
}

This paper proposes a platform for portal and local repositories. Our methodology aims not only at construction of portal site but also at supporting capture of digital contents transformed from interview videos with intellectuals.

@inbook{
 type = {inbook},
 year = {2005},
 pages = {493-495},
 volume = {LNCS 3652},
 websites = {http://link.springer.com/10.1007/11551362_51},
 publisher = {Springer},
 city = {Berlin, Heidelberg},
 id = {a100116b-73c2-3090-a0a3-1d246c461e13},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T02:59:57.734Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 bibtype = {inbook},
 author = {Takasu, Atsuhiro and Aihara, Kenro},
 doi = {10.1007/11551362_51},
 chapter = {An Effective Access Mechanism to Digital Interview Archives},
 title = {Research and Advanced Technology for Digital Libraries. ECDL 2005}
}

@article{
 title = {Bibliographic component extraction from references based on a text recognition error model},
 type = {article},
 year = {2005},
 keywords = {[Bibliographic matching, Digital library, Hidden M},
 volume = {36},
 id = {04f34ce7-b905-3caf-aa10-34504fc60dad},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Citation linkage is important for information retrieval and navigation in digital libraries. This paper proposes a novel method to extract bibliographic components from reference strings obtained from scanned document images. The proposed method uses an extended hidden Markov model representing both OCR error patterns and the syntactical structure of references simultaneously. Rule-based systems usually have difficulty in obtaining rules. The proposed method gives a solution to this problem by parameter estimation of the model from training data. We applied the proposed method to references obtained by OCR with a large bibliographic database, and achieved about 90% extraction accuracy. Furthermore, the error analysis shows that the proposed method can find the approximate position of bibliographic components in the reference with very high accuracy. © 2005 Wiley Periodicals, Inc.},
 bibtype = {article},
 author = {Takasu, A. and Aihara, K.},
 doi = {10.1002/scj.20323},
 journal = {Systems and Computers in Japan},
 number = {7}
}

Citation linkage is important for information retrieval and navigation in digital libraries. This paper proposes a novel method to extract bibliographic components from reference strings obtained from scanned document images. The proposed method uses an extended hidden Markov model representing both OCR error patterns and the syntactical structure of references simultaneously. Rule-based systems usually have difficulty in obtaining rules. The proposed method gives a solution to this problem by parameter estimation of the model from training data. We applied the proposed method to references obtained by OCR with a large bibliographic database, and achieved about 90% extraction accuracy. Furthermore, the error analysis shows that the proposed method can find the approximate position of bibliographic components in the reference with very high accuracy. © 2005 Wiley Periodicals, Inc.

@inproceedings{
 title = {The improvement in unstructured peer-to-peer networks using small-world model},
 type = {inproceedings},
 year = {2005},
 keywords = {[Indexing, Peer-to-peer, Query processing, Routing},
 volume = {8},
 id = {2336ccaa-b154-326a-a023-4258e5f189e8},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Due to effective data sharing for open cooperative working environments on unstructured peer-to-peer (P2P) systems, one of the issues is the query processing performance. In this paper we study the effects of P2P search performance using a distributed indexing mechanism, Direct Indices (DIs) which are the key to reducing the number of flooding queries, and improving scalability over P2P networks. On P2P systems using DIs, a peer enables to construct edges to others directly. This network model is based on the small-world models, which have two characteristic parameters, a small average path length and a large clustering coefficient. We show that our P2P system with DIs is approximated to the small-world phenomenon via simulations, and that query processing on an unstructured P2P system is affected by the power of the small-world phenomenon.},
 bibtype = {inproceedings},
 author = {Yamada, T. and Aihara, K. and Takasu, A. and Adachi, J.},
 booktitle = {WMSCI 2005 - The 9th World Multi-Conference on Systemics, Cybernetics and Informatics, Proceedings}
}

Due to effective data sharing for open cooperative working environments on unstructured peer-to-peer (P2P) systems, one of the issues is the query processing performance. In this paper we study the effects of P2P search performance using a distributed indexing mechanism, Direct Indices (DIs) which are the key to reducing the number of flooding queries, and improving scalability over P2P networks. On P2P systems using DIs, a peer enables to construct edges to others directly. This network model is based on the small-world models, which have two characteristic parameters, a small average path length and a large clustering coefficient. We show that our P2P system with DIs is approximated to the small-world phenomenon via simulations, and that query processing on an unstructured P2P system is affected by the power of the small-world phenomenon.

@inproceedings{
 title = {Adaptive replication method based on peer behavior pattern in unstructured peer-to-peer systems},
 type = {inproceedings},
 year = {2005},
 volume = {2005},
 id = {c1963ba2-942f-3b74-aa0d-863c43b39014},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {In this paper, we propose replication methods to perform effective document sharing on a peer-to-peer(P2P) system where peers frequently join and leave the system. The proposed method uses the relevancy and usefulness of peers to determine how many replications should be made, and where to locate these replications. This paper shows empirically that the proposed method improves the efficiency of query processing. © 2005 IEEE.},
 bibtype = {inproceedings},
 author = {Yamada, T. and Aihara, K. and Takasu, A. and Adachi, J.},
 doi = {10.1109/ICDE.2005.182},
 booktitle = {Proceedings - International Workshop on Biomedical Data Engineering, BMDE2005}
}

In this paper, we propose replication methods to perform effective document sharing on a peer-to-peer(P2P) system where peers frequently join and leave the system. The proposed method uses the relevancy and usefulness of peers to determine how many replications should be made, and where to locate these replications. This paper shows empirically that the proposed method improves the efficiency of query processing. © 2005 IEEE.

@inproceedings{
 title = {A Topic-Based Index Mechanism Using Usefulness of Peers in Unstructured Peer-to-Peer Networks},
 type = {inproceedings},
 year = {2005},
 pages = {134-139},
 month = {2},
 city = {Innsbruck, Austria},
 id = {0a7dd8dd-25ce-3bb8-879b-fabe88966fc0},
 created = {2025-08-03T13:12:14.870Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:12:28.231Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Yamada, Taizo and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 booktitle = {The Twenty-Third IASTED International Conference on Databases and Applications (DBA2005)}
}

@inproceedings{
 title = {An index system for dynamic peer-to-peer network based on document usefulness},
 type = {inproceedings},
 year = {2004},
 volume = {15},
 id = {9b988ffd-03d8-38f7-9c14-f4b2428f374d},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:13:13.513Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Peer-to-Peer (P2P) systems are applied to cooperative working environments, but require more effective document sharing mechanisms, particularly, more efficient query processing. To achieve efficient query processing in the environment, we study a distributed indexing mechanism, normal Direct Indices (Dis), which are the key to reducing the number of queries and improving scalability over P2P networks. However, query processing with DIs cannot efficiently deal with an increase in the number of document categories because a DI holds the peer's information, not the document information. Therefore, we propose an unproved DI, query-based DIs (QDIs). We assume that the peer is relevant to the query terms that the peer sends. This assumption allows each peer with correct QDIs to forward the query messages to their destination. To validate the QDIs, we show experimental results on query processing with and without DIs and show the improvement in network performance.},
 bibtype = {inproceedings},
 author = {Yamada, Taizo and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 booktitle = {International Conference on Database and Expert Systems Applications - DEXA}
}

Peer-to-Peer (P2P) systems are applied to cooperative working environments, but require more effective document sharing mechanisms, particularly, more efficient query processing. To achieve efficient query processing in the environment, we study a distributed indexing mechanism, normal Direct Indices (Dis), which are the key to reducing the number of queries and improving scalability over P2P networks. However, query processing with DIs cannot efficiently deal with an increase in the number of document categories because a DI holds the peer's information, not the document information. Therefore, we propose an unproved DI, query-based DIs (QDIs). We assume that the peer is relevant to the query terms that the peer sends. This assumption allows each peer with correct QDIs to forward the query messages to their destination. To validate the QDIs, we show experimental results on query processing with and without DIs and show the improvement in network performance.

@inproceedings{
 title = {Replica placement for effective document sharing mechanisms in peer-to-peer networks},
 type = {inproceedings},
 year = {2004},
 keywords = {[Load-balancing, Peer-to-peer, Query processing, R},
 id = {05cc5bf1-d230-3b14-810f-ec6ea29f16e4},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:13:29.865Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Recently, peer-to-peer (P2P) systems have become popular for both data exchange and cooperative working environments. However, an effective document sharing mechanism is required to work cooperatively on P2P systems. In this paper, we propose replication methods to perform effective document sharing on a P2P system where peers frequently join and leave the system. The proposed method uses the relevancy and usefulness of peers to determine how many replications should be made, and where to locate these replications. In addition, we propose an update procedure for replicated documents. This paper shows empirically that the proposed method improves the efficiency of query processing and load-balancing.},
 bibtype = {inproceedings},
 author = {Yamada, Taizo and Aihara, Kenro and Takasu, Atsuhiro and Adachi, Jun},
 booktitle = {Proceedings of the Eighth IASTED International Conference on Internet and Multimedia Systems and Applications}
}

Recently, peer-to-peer (P2P) systems have become popular for both data exchange and cooperative working environments. However, an effective document sharing mechanism is required to work cooperatively on P2P systems. In this paper, we propose replication methods to perform effective document sharing on a P2P system where peers frequently join and leave the system. The proposed method uses the relevancy and usefulness of peers to determine how many replications should be made, and where to locate these replications. In addition, we propose an update procedure for replicated documents. This paper shows empirically that the proposed method improves the efficiency of query processing and load-balancing.

@article{
 title = {テキスト認識エラーモデルによる引用文献文字列からの書誌要素の抽出},
 type = {article},
 year = {2004},
 pages = {1298-1308},
 volume = {87},
 id = {d9746a4d-a088-3187-b50b-a957387da72a},
 created = {2021-10-12T05:26:23.778Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:01:34.800Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {電子図書館や文献検索において引用関係などに基づいて文献相互のリンクを形成することは，情報検索やナビゲーションを行う上で重要である．本論文では，電子化される以前の文献も含めた相互リンクを実現するために，文書画像解析の結果得られた引用文献文字列から書誌要素を抽出する方法を提案する．提案手法では，引用文献文字列の構文構造とOCRの認識誤りを扱うために，隠れマルコフモデルを拡張した統計モデルを用いる．ルールベースのシステムでは，ルールの獲得が大きな問題となるが，提案手法では，訓練データを用いた統計モデルのパラメータ推定アルゴリズムによってモデルを低コストで獲得できる．本論文では，OCRによって得られた引用文献文字列と大規模書誌データベースを用いて，提案手法の評価を行い，90％程度の精度で書誌要素を抽出できることを示す．更に，書誌要素の抽出に失敗した場合でも，引用文献文字列中の書誌要素のおおよその位置を特定できることを示し，文献の相互リンクのための書誌マッチングにおいて，高い精度で書誌要素を抽出できることを実験的に示す．},
 bibtype = {article},
 author = {高須, 淳宏 and 相原, 健郎},
 journal = {電子情報通信学会論文誌 D-II, 情報・システム, II-パターン処理},
 number = {6}
}

電子図書館や文献検索において引用関係などに基づいて文献相互のリンクを形成することは，情報検索やナビゲーションを行う上で重要である．本論文では，電子化される以前の文献も含めた相互リンクを実現するために，文書画像解析の結果得られた引用文献文字列から書誌要素を抽出する方法を提案する．提案手法では，引用文献文字列の構文構造とOCRの認識誤りを扱うために，隠れマルコフモデルを拡張した統計モデルを用いる．ルールベースのシステムでは，ルールの獲得が大きな問題となるが，提案手法では，訓練データを用いた統計モデルのパラメータ推定アルゴリズムによってモデルを低コストで獲得できる．本論文では，OCRによって得られた引用文献文字列と大規模書誌データベースを用いて，提案手法の評価を行い，90％程度の精度で書誌要素を抽出できることを示す．更に，書誌要素の抽出に失敗した場合でも，引用文献文字列中の書誌要素のおおよその位置を特定できることを示し，文献の相互リンクのための書誌マッチングにおいて，高い精度で書誌要素を抽出できることを実験的に示す．

@article{
 title = {Peer-to-Peerシステム上での効率的なデータ配置による問い合わせ処理とロードバランスへの寄与},
 type = {article},
 year = {2004},
 pages = {93-101},
 volume = {45},
 id = {6c2c9bd8-e663-339c-bf4f-294c0d63850e},
 created = {2025-08-03T13:15:51.909Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:03:39.524Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {現在，P2P システムは一般的に認知されはじめ，単なるデータ交換システムだけではなく，協調作業のシステムとしても注目を浴びている．しかしながら，協調作業を P2P システム上で行うためには，効果的なデータ共有を行う環境整備が求められている．本研究ではシステム上の各ピアの接続状況が激しく変動するシステムを仮定し，そこでの P2P システム上での効果的なデータ共有を行うための効率的なデータ配置方法を考案している．ここでの提案方式では，問合せ処理およびロード バランシングの向上を期待した方式である．またシステム上の各データの有用性によって複製対象データを決定し，複製されるデータ数を決定する方法も提案する．さらに，データの更新時のデータ再配置に関しても考慮している．提案方式の効果を確かめ，しかもデータ配置のシステムへの寄与の状況を確かめるべく，各種実験を行っている．これらの実験では問合せ処理およびロード バランシングの状況を計測し，効果的な問合せ処理およびロード バランシングを行うことが可能であることが示す．

Recently P2P systems have been popular as not only mere data changing systems but also cooperation working systems. However, to do the cooperation working on P2P systems, effec-
tive data sharing systems are requested. In the paper, we assume systems that a condition of connection of each peer on systems changes violently, and propose techniques of efficient data placement to do effective data sharing on P2P systems. The techniques which we propose are expected that query processing and load-balancing are improved. Moreover, we propose a technique which a usefulness of data a peer hold defines replicated data, and a usefulness of the peer gives the definition of the number of replicas with which each peer can be generated.
In addition, we consider data replacement when each data on systems is updated. To ensure effectiveness of the proposal techniques, we show experimental results on query processing and load-balancing.},
 bibtype = {article},
 author = {山田, 太造 and 相原, 健郎 and 高須, 淳宏 and 安達, 淳},
 journal = {情報処理学会論文誌：データベース},
 number = {7}
}

現在，P2P システムは一般的に認知されはじめ，単なるデータ交換システムだけではなく，協調作業のシステムとしても注目を浴びている．しかしながら，協調作業を P2P システム上で行うためには，効果的なデータ共有を行う環境整備が求められている．本研究ではシステム上の各ピアの接続状況が激しく変動するシステムを仮定し，そこでの P2P システム上での効果的なデータ共有を行うための効率的なデータ配置方法を考案している．ここでの提案方式では，問合せ処理およびロード バランシングの向上を期待した方式である．またシステム上の各データの有用性によって複製対象データを決定し，複製されるデータ数を決定する方法も提案する．さらに，データの更新時のデータ再配置に関しても考慮している．提案方式の効果を確かめ，しかもデータ配置のシステムへの寄与の状況を確かめるべく，各種実験を行っている．これらの実験では問合せ処理およびロード バランシングの状況を計測し，効果的な問合せ処理およびロード バランシングを行うことが可能であることが示す． Recently P2P systems have been popular as not only mere data changing systems but also cooperation working systems. However, to do the cooperation working on P2P systems, effec- tive data sharing systems are requested. In the paper, we assume systems that a condition of connection of each peer on systems changes violently, and propose techniques of efficient data placement to do effective data sharing on P2P systems. The techniques which we propose are expected that query processing and load-balancing are improved. Moreover, we propose a technique which a usefulness of data a peer hold defines replicated data, and a usefulness of the peer gives the definition of the number of replicas with which each peer can be generated. In addition, we consider data replacement when each data on systems is updated. To ensure effectiveness of the proposal techniques, we show experimental results on query processing and load-balancing.

@article{
 title = {テキスト認識エラーモデルによる引用文献文字列からの書誌要素の抽出},
 type = {article},
 year = {2004},
 pages = {1298-1308},
 volume = {87},
 id = {57fceac0-dc87-398d-83a1-61cad5dd7dc7},
 created = {2025-08-03T13:17:54.327Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:19:00.101Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {article},
 author = {高須, 淳宏 and 相原, 健郎},
 journal = {電子情報通信学会論文誌 D-II, 情報・システム, II-パターン処理},
 number = {6}
}

@article{
 title = {Experimental study on the relationship between the performance and training data size of text categorization},
 type = {article},
 year = {2003},
 keywords = {[Categorization Performance, Decision Tree, Suppor},
 id = {6085523a-75b9-3798-a196-816b390c9de7},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {This paper discusses the relationship between the performance of text categorization and the size of training data. In the studies of text categorization, a pair of training data and test data is used to evaluate the performance of text categorizers. In this paper we conduct experiments on the convergence of text categorizers with respect to the training data size and discuss the validity of text categorization experiments.},
 bibtype = {article},
 author = {Takasu, A. and Aihara, K.},
 journal = {NII Journal},
 number = {6}
}

This paper discusses the relationship between the performance of text categorization and the size of training data. In the studies of text categorization, a pair of training data and test data is used to evaluate the performance of text categorizers. In this paper we conduct experiments on the convergence of text categorizers with respect to the training data size and discuss the validity of text categorization experiments.

@article{
 title = {Domain visualization based on authorized documents},
 type = {article},
 year = {2003},
 keywords = {[Human-Computer Interaction, Information Retrieval},
 id = {34d109c1-e480-32e2-a327-27b69ba8b7db},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {In this paper, we show our visualization approach for supporting interactive document retrieval based on authorized documents, such as survey articles or domain dictionaries. We assume that it is effective for accessing the required information resources that the system let users recognize the information space inclusively. Our research, therefore, aims to support users to access the information resources, which the user require, through the interaction between visualized information space and the user. We will use the term "domain visualization" to refer to visualizing the structure based on contents of the domain. We use Self-Organizing Maps to classify documents and visualize the domain. We show an effectiveness of using domain dictionaries as feature of documents and survey articles as initial sets of SOM learning.},
 bibtype = {article},
 author = {Aihara, K. and Takasu, A.},
 journal = {NII Journal},
 number = {5}
}

In this paper, we show our visualization approach for supporting interactive document retrieval based on authorized documents, such as survey articles or domain dictionaries. We assume that it is effective for accessing the required information resources that the system let users recognize the information space inclusively. Our research, therefore, aims to support users to access the information resources, which the user require, through the interaction between visualized information space and the user. We will use the term "domain visualization" to refer to visualizing the structure based on contents of the domain. We use Self-Organizing Maps to classify documents and visualize the domain. We show an effectiveness of using domain dictionaries as feature of documents and survey articles as initial sets of SOM learning.

@inproceedings{
 title = {A Distributed Index System for Efficient Query Processing in Peer-to-Peer Networks},
 type = {inproceedings},
 year = {2003},
 volume = {I},
 id = {75dde2ee-504c-308c-82e2-775aa43c06d0},
 created = {2016-11-25T15:05:19.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {Peer-to-Peer (P2P) systems have the advantage of high autonomy. However, query processing in P2P system is inefficient because query messages propagate in the P2P network. To realize efficient and scalable systems without losing the autonomy of the P2P system, this paper proposes an index system called Direct Indices (DIs) in which the indexes keep information about the usefulness of peers and queries are forwarded to useful peers. According to the query result, peers access the document directly. In this way, bandwidth consumption is reduced and scalability of P2P networks is achieved. This paper shows the performance of Dis experimentally.},
 bibtype = {inproceedings},
 author = {Yamada, T. and Aihara, K. and Takasu, A. and Adachi, J.},
 booktitle = {IEEE Pacific RIM Conference on Communications, Computers, and Signal Processing - Proceedings}
}

Peer-to-Peer (P2P) systems have the advantage of high autonomy. However, query processing in P2P system is inefficient because query messages propagate in the P2P network. To realize efficient and scalable systems without losing the autonomy of the P2P system, this paper proposes an index system called Direct Indices (DIs) in which the indexes keep information about the usefulness of peers and queries are forwarded to useful peers. According to the query result, peers access the document directly. In this way, bandwidth consumption is reduced and scalability of P2P networks is achieved. This paper shows the performance of Dis experimentally.

@inproceedings{
 title = {DVHMM: variable length text recognition error model},
 type = {inproceedings},
 year = {2002},
 pages = {110-114},
 volume = {3},
 issue = {3},
 websites = {http://ieeexplore.ieee.org/document/1047807/},
 publisher = {IEEE},
 city = {Quebec City},
 id = {58d3febc-9556-30ec-a4c5-cd9456e7c3bf},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:36:52.747Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {This paper proposes a text recognition error model called the dual variable length output hidden Markov model (DVHMM) and gives a parameter estimation algorithm based on the EM algorithm. Although existing probabilistic error models are limited to substitution (1,1), insertion (1,0), and deletion (0,1) errors, the DVHMM can handle error patterns of any pair (i, j) of lengths including substitution, insertion, and deletion. © 2002 IEEE.},
 bibtype = {inproceedings},
 author = {Takasu, Atsuhiro and Aihara, Kenro},
 doi = {10.1109/ICPR.2002.1047807},
 booktitle = {2002 International Conference on Pattern Recognition}
}

This paper proposes a text recognition error model called the dual variable length output hidden Markov model (DVHMM) and gives a parameter estimation algorithm based on the EM algorithm. Although existing probabilistic error models are limited to substitution (1,1), insertion (1,0), and deletion (0,1) errors, the DVHMM can handle error patterns of any pair (i, j) of lengths including substitution, insertion, and deletion. © 2002 IEEE.

@inproceedings{
 title = {Supporting Collaborative Development and Operation Based on Integrated Heterogeneous Data with Temporal Information},
 type = {inproceedings},
 year = {2002},
 pages = {63-64},
 id = {2cc339e3-1f56-3aae-961a-ca7771b75d37},
 created = {2025-08-03T13:20:43.997Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-03T13:20:47.288Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 bibtype = {inproceedings},
 author = {Aihara, Kenro and Takasu, Atsuhiro and Kato, Yoshikiyo},
 booktitle = {Conference Supplement of ACM 2002 Conference on Computer Supported Cooperative Work}
}

@inbook{
 type = {inbook},
 year = {2001},
 keywords = {[Customization, Human-Computer Interaction, Inform},
 pages = {207-209},
 volume = {LNCS 2109},
 websites = {http://link.springer.com/10.1007/3-540-44566-8_22},
 publisher = {Springer},
 city = {Berlin, Heidelberg},
 id = {24c3c886-a31d-32e5-8ee0-878d7efc3909},
 created = {2016-11-25T15:05:17.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:28:34.050Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {© Springer-Verlag Berlin Heidelberg 2001.This paper proposes an customization technique to supporting interactive document retrieval in unorganized open information space like WWW. We assume that taxonomical thought is one of the most important and skilled operations for us when we organize or store information. The proposed methodology, therefore, handles hierarchical categories of documents. The system can be customized through users’ modification of categories. The features of the proposed approach are (1) visualization of document categories for interaction, (2) initialization of categories by hierarchical clustering method, (3) customization of categories by support vector machine techniques, (4) additional attributes for individual implicit cognitive aspects.},
 bibtype = {inbook},
 author = {Aihara, Kenro and Takasu, Atsuhiro},
 doi = {10.1007/3-540-44566-8_22},
 chapter = {Category Based Customization Approach for Information Retrieval},
 title = {User Modeling 2001. UM 2001}
}

© Springer-Verlag Berlin Heidelberg 2001.This paper proposes an customization technique to supporting interactive document retrieval in unorganized open information space like WWW. We assume that taxonomical thought is one of the most important and skilled operations for us when we organize or store information. The proposed methodology, therefore, handles hierarchical categories of documents. The system can be customized through users’ modification of categories. The features of the proposed approach are (1) visualization of document categories for interaction, (2) initialization of categories by hierarchical clustering method, (3) customization of categories by support vector machine techniques, (4) additional attributes for individual implicit cognitive aspects.

@article{
 title = {記憶の想起に基づく創造性支援},
 type = {article},
 year = {2001},
 pages = {1377-1386},
 volume = {42},
 websites = {https://ipsj.ixsq.nii.ac.jp/records/11934},
 month = {6},
 id = {fff1d168-72fe-3192-bfb7-d9ef0fbe76c4},
 created = {2025-08-04T01:07:39.136Z},
 accessed = {2025-08-04},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:09:56.313Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 citation_key = {Aihara:IPSJ:2001},
 private_publication = {false},
 abstract = {本論文は，人間の創造的な思考を支援するための手法と，それに基づく支援システムを提案する．まず，創造性についての認知科学研究に基づいた思考過程のモデルを提示し，思考の制約となっているものを変更することで創造的な思考が進むことを述べる．そして，思考の制約のうち，記憶の想起の障害となっている時間的な制約に着目し，それをシステムによってユーザに変更を促すことで創造的な思考を支援することを提案している．提案手法では，時間的な制約を空間的な制約に変換してユーザに提示することで，ユーザの想起を促し，それによって思考を支援することを行う．支援システムEn Passant 2は，思考活動で最もよく用いられる紙に書かれたメモを利用し，それをシステムに入力することで，ユーザの記憶の想起のきっかけを与えるのに利用している．ユーザは，システム上でメモのマークをつけることでメモを特徴づける．それを用いてシステムはメモ間の関連度を算出し，メモを2次元空間上にマップしてユーザに提示する．システムを用いた長期実験を行い，システムの有用性を示している．},
 bibtype = {article},
 author = {相原, 健郎 and 堀, 浩一},
 journal = {情報処理学会誌},
 number = {6}
}

本論文は，人間の創造的な思考を支援するための手法と，それに基づく支援システムを提案する．まず，創造性についての認知科学研究に基づいた思考過程のモデルを提示し，思考の制約となっているものを変更することで創造的な思考が進むことを述べる．そして，思考の制約のうち，記憶の想起の障害となっている時間的な制約に着目し，それをシステムによってユーザに変更を促すことで創造的な思考を支援することを提案している．提案手法では，時間的な制約を空間的な制約に変換してユーザに提示することで，ユーザの想起を促し，それによって思考を支援することを行う．支援システムEn Passant 2は，思考活動で最もよく用いられる紙に書かれたメモを利用し，それをシステムに入力することで，ユーザの記憶の想起のきっかけを与えるのに利用している．ユーザは，システム上でメモのマークをつけることでメモを特徴づける．それを用いてシステムはメモ間の関連度を算出し，メモを2次元空間上にマップしてユーザに提示する．システムを用いた長期実験を行い，システムの有用性を示している．

@article{
 title = {Variance based classifier comparison in text categorization},
 type = {article},
 year = {2000},
 id = {85b07ec6-a639-31d1-bd11-b80e2ce8f167},
 created = {2016-11-25T15:05:18.000Z},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2017-03-28T14:06:45.562Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {The selection of a classifier from various classification methods and its affect on the size of training data was discussed. The variance of classifier was evaluated for assessing the reliability of classification method. Each document was represented with a feature vector consisting of weighted frequency terms in the text categorization process.},
 bibtype = {article},
 author = {Takasu, Atsuhiro and Aihara, Kenro},
 journal = {SIGIR Forum (ACM Special Interest Group on Information Retrieval)}
}

The selection of a classifier from various classification methods and its affect on the size of training data was discussed. The variance of classifier was evaluated for assessing the reliability of classification method. Each document was represented with a feature vector consisting of weighted frequency terms in the text categorization process.

@article{
 title = {Enhancing creativity through reorganising mental space concealed in a research notes stack},
 type = {article},
 year = {1998},
 keywords = {Change of constraint,Creativity support system,Naturalness of media},
 pages = {469-478},
 volume = {11},
 month = {12},
 publisher = {Elsevier},
 day = {1},
 id = {45010671-0d90-3cde-b13e-3403a5f1ad7c},
 created = {2025-08-04T01:12:15.502Z},
 accessed = {2025-08-04},
 file_attached = {false},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T01:15:01.245Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {In this article, we propose a method and its implementation to aid the process of creative thinking. The objective of this research is to enhance human scientific creativity with computers. In order to aid the process of creative thinking, we have implemented a system named En Passant 2, which stores the user's research notes and gives the triggers to recall his/her memories in current context. En Passant 2 has a function to deal with indices and a time attribute of the user's thought explicitly, and shows the user's notepads related to his/her awareness of the issues. © 1998 Elsevier Science B.V. All rights reserved.},
 bibtype = {article},
 author = {Aihara, Kenro and Hori, Koichi},
 doi = {10.1016/S0950-7051(98)00080-X},
 journal = {Knowledge-Based Systems},
 number = {7-8}
}

In this article, we propose a method and its implementation to aid the process of creative thinking. The objective of this research is to enhance human scientific creativity with computers. In order to aid the process of creative thinking, we have implemented a system named En Passant 2, which stores the user's research notes and gives the triggers to recall his/her memories in current context. En Passant 2 has a function to deal with indices and a time attribute of the user's thought explicitly, and shows the user's notepads related to his/her awareness of the issues. © 1998 Elsevier Science B.V. All rights reserved.

@article{
 title = {断片的な情報の集まりから知識を構築する過程の支援},
 type = {article},
 year = {1996},
 pages = {432-439},
 volume = {11},
 id = {69826a64-f6ea-31c2-a594-f8a0acd15a7b},
 created = {2025-08-03T13:23:25.826Z},
 file_attached = {true},
 profile_id = {c36e54e0-03a7-3bfb-870f-0468e54a2cc1},
 last_modified = {2025-08-04T03:06:04.226Z},
 read = {true},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {true},
 private_publication = {false},
 abstract = {In this paper, we present a method and its implementation to aid the process of building knowledge out of chunks of information. The chunks of information, which can be the source of new knowledge, often exist in so called tacit dimension. This paper deals with these unclassified chunks of information that haven't been systematized yet but has potential to be systematized. In order to aid the process of building knowledge out of those chunks, we give a system which maps the chunks into a metric space. Through the reconfiguration of this space, the user can gradually structurize the chunks of information into systematized knowledge. The unique feature of the proposed system compared with the existing ones lies in the function to deal with the users point of view. The system reveals the user's hidden point of view explicitly, and this helps the process of systematization of knowledge much. We have implemented a system CASSIS, which stands for Computer-ASSIsted Systematization, and have carried out several experiments. We show the configuration of the system and the results of the experiments.},
 bibtype = {article},
 author = {相原, 健郎 and 堀, 浩一 and 大須賀, 節雄},
 journal = {人工知能学会誌},
 number = {3}
}

In this paper, we present a method and its implementation to aid the process of building knowledge out of chunks of information. The chunks of information, which can be the source of new knowledge, often exist in so called tacit dimension. This paper deals with these unclassified chunks of information that haven't been systematized yet but has potential to be systematized. In order to aid the process of building knowledge out of those chunks, we give a system which maps the chunks into a metric space. Through the reconfiguration of this space, the user can gradually structurize the chunks of information into systematized knowledge. The unique feature of the proposed system compared with the existing ones lies in the function to deal with the users point of view. The system reveals the user's hidden point of view explicitly, and this helps the process of systematization of knowledge much. We have implemented a system CASSIS, which stands for Computer-ASSIsted Systematization, and have carried out several experiments. We show the configuration of the system and the results of the experiments.