Content-based artwork recommendation: integrating painting metadata with neural and manually-engineered visual features. Messina, P., Dominguez, V., Parra, D., Trattner, C., & Soto, A. User Modeling and User-Adapted Interaction, 2019. Paper abstract bibtex 1 download Recommender Systems help us deal with information overload by suggesting relevant items based on our personal preferences. Although there is a large body of research in areas such as movies or music, artwork recommendation has received comparatively little attention, despite the continuous growth of the artwork market. Most previous research has relied on ratings and metadata, and a few recent works have exploited visual features extracted with deep neural networks (DNN) to recommend digital art. In this work, we contribute to the area of content-based artwork recommendation of physical paintings by studying the impact of the aforementioned features (artwork metadata, neural visual features), as well as manually-engineered visual features, such as naturalness, brightness and contrast. We implement and evaluate our method using transactional data from UGallery.com, an online artwork store. Our results show that artwork recommendations based on a hybrid combination of artist preference, curated attributes, deep neural visual features and manually-engineered visual features produce the best performance. Moreover, we discuss the trade-off between automatically obtained DNN features and manually-engineered visual features for the purpose of explainability, as well as the impact of user profile size on predictions. Our research informs the development of next-generation content-based artwork recommenders which rely on different types of data, from text to multimedia.
@Article{ messina:etal:2019,
author = {Pablo Messina and Vicente Dominguez and Denis Parra and
Christoph Trattner and Alvaro Soto},
title = {Content-based artwork recommendation: integrating painting
metadata with neural and manually-engineered visual
features},
journal = {User Modeling and User-Adapted Interaction},
volume = {29},
number = {2},
year = {2019},
abstract = {Recommender Systems help us deal with information overload
by suggesting relevant items based on our personal
preferences. Although there is a large body of research in
areas such as movies or music, artwork recommendation has
received comparatively little attention, despite the
continuous growth of the artwork market. Most previous
research has relied on ratings and metadata, and a few
recent works have exploited visual features extracted with
deep neural networks (DNN) to recommend digital art. In
this work, we contribute to the area of content-based
artwork recommendation of physical paintings by studying
the impact of the aforementioned features (artwork
metadata, neural visual features), as well as
manually-engineered visual features, such as naturalness,
brightness and contrast. We implement and evaluate our
method using transactional data from UGallery.com, an
online artwork store. Our results show that artwork
recommendations based on a hybrid combination of artist
preference, curated attributes, deep neural visual features
and manually-engineered visual features produce the best
performance. Moreover, we discuss the trade-off between
automatically obtained DNN features and manually-engineered
visual features for the purpose of explainability, as well
as the impact of user profile size on predictions. Our
research informs the development of next-generation
content-based artwork recommenders which rely on different
types of data, from text to multimedia.},
url = {https://link.springer.com/article/10.1007/s11257-018-9206-9}
}
Downloads: 1
{"_id":"efSZ5mCwkM437htCc","bibbaseid":"messina-dominguez-parra-trattner-soto-contentbasedartworkrecommendationintegratingpaintingmetadatawithneuralandmanuallyengineeredvisualfeatures-2019","authorIDs":["32ZR23o2BFySHbtQK","3ear6KFZSRqbj6YeT","4Pq6KLaQ8jKGXHZWH","54578d9a2abc8e9f370004f0","5e126ca5a4cabfdf01000053","5e158f76f1f31adf01000118","5e16174bf67f7dde010003ad","5e1f631ae8f5ddde010000eb","5e1f7182e8f5ddde010001ff","5e26da3642065ede01000066","5e3acefaf2a00cdf010001c8","5e62c3aecb259cde010000f9","5e65830c6e5f4cf3010000e7","5e666dfc46e828de010002c9","6cMBYieMJhf6Nd58M","6w6sGsxYSK2Quk6yZ","7xDcntrrtC62vkWM5","ARw5ReidxxZii9TTZ","DQ4JRTTWkvKXtCNCp","GbYBJvxugXMriQwbi","HhRoRmBvwWfD4oLyK","JFk6x26H6LZMoht2n","JvArGGu5qM6EvSCvB","LpqQBhFH3PxepH9KY","MT4TkSGzAp69M3dGt","QFECgvB5v2i4j2Qzs","RKv56Kes3h6FwEa55","Rb9TkQ3KkhGAaNyXq","RdND8NxcJDsyZdkcK","SpKJ5YujbHKZnHc4v","TSRdcx4bbYKqcGbDg","W8ogS2GJa6sQKy26c","WTi3X2fT8dzBN5d8b","WfZbctNQYDBaiYW6n","XZny8xuqwfoxzhBCB","Xk2Q5qedS5MFHvjEW","bbARiTJLYS79ZMFbk","cBxsyeZ37EucQeBYK","cFyFQps7W3Sa2Wope","dGRBfr8zhMmbwK6eP","eRLgwkrEk7T7Lmzmf","fMYSCX8RMZap548vv","g6iKCQCFnJgKYYHaP","h2hTcQYuf2PB3oF8t","h83jBvZYJPJGutQrs","jAtuJBcGhng4Lq2Nd","pMoo2gotJcdDPwfrw","q5Zunk5Y2ruhw5vyq","rzNGhqxkbt2MvGY29","uC8ATA8AfngWpYLBq","uoJ7BKv28Q6TtPmPp","vMiJzqEKCsBxBEa3v","vQE6iTPpjxpuLip2Z","wQDRsDjhgpMJDGxWX","wbNg79jvDpzX9zHLK","wk86BgRiooBjy323E","zCbPxKnQGgDHiHMWn","zf9HENjsAzdWLMDAu"],"author_short":["Messina, P.","Dominguez, V.","Parra, D.","Trattner, C.","Soto, A."],"bibdata":{"bibtype":"article","type":"article","author":[{"firstnames":["Pablo"],"propositions":[],"lastnames":["Messina"],"suffixes":[]},{"firstnames":["Vicente"],"propositions":[],"lastnames":["Dominguez"],"suffixes":[]},{"firstnames":["Denis"],"propositions":[],"lastnames":["Parra"],"suffixes":[]},{"firstnames":["Christoph"],"propositions":[],"lastnames":["Trattner"],"suffixes":[]},{"firstnames":["Alvaro"],"propositions":[],"lastnames":["Soto"],"suffixes":[]}],"title":"Content-based artwork recommendation: integrating painting metadata with neural and manually-engineered visual features","journal":"User Modeling and User-Adapted Interaction","volume":"29","number":"2","year":"2019","abstract":"Recommender Systems help us deal with information overload by suggesting relevant items based on our personal preferences. Although there is a large body of research in areas such as movies or music, artwork recommendation has received comparatively little attention, despite the continuous growth of the artwork market. Most previous research has relied on ratings and metadata, and a few recent works have exploited visual features extracted with deep neural networks (DNN) to recommend digital art. In this work, we contribute to the area of content-based artwork recommendation of physical paintings by studying the impact of the aforementioned features (artwork metadata, neural visual features), as well as manually-engineered visual features, such as naturalness, brightness and contrast. We implement and evaluate our method using transactional data from UGallery.com, an online artwork store. Our results show that artwork recommendations based on a hybrid combination of artist preference, curated attributes, deep neural visual features and manually-engineered visual features produce the best performance. Moreover, we discuss the trade-off between automatically obtained DNN features and manually-engineered visual features for the purpose of explainability, as well as the impact of user profile size on predictions. Our research informs the development of next-generation content-based artwork recommenders which rely on different types of data, from text to multimedia.","url":"https://link.springer.com/article/10.1007/s11257-018-9206-9","bibtex":"@Article{\t messina:etal:2019,\n author\t= {Pablo Messina and Vicente Dominguez and Denis Parra and\n\t\t Christoph Trattner and Alvaro Soto},\n title\t\t= {Content-based artwork recommendation: integrating painting\n\t\t metadata with neural and manually-engineered visual\n\t\t features},\n journal\t= {User Modeling and User-Adapted Interaction},\n volume\t= {29},\n number\t= {2},\n year\t\t= {2019},\n abstract\t= {Recommender Systems help us deal with information overload\n\t\t by suggesting relevant items based on our personal\n\t\t preferences. Although there is a large body of research in\n\t\t areas such as movies or music, artwork recommendation has\n\t\t received comparatively little attention, despite the\n\t\t continuous growth of the artwork market. Most previous\n\t\t research has relied on ratings and metadata, and a few\n\t\t recent works have exploited visual features extracted with\n\t\t deep neural networks (DNN) to recommend digital art. In\n\t\t this work, we contribute to the area of content-based\n\t\t artwork recommendation of physical paintings by studying\n\t\t the impact of the aforementioned features (artwork\n\t\t metadata, neural visual features), as well as\n\t\t manually-engineered visual features, such as naturalness,\n\t\t brightness and contrast. We implement and evaluate our\n\t\t method using transactional data from UGallery.com, an\n\t\t online artwork store. Our results show that artwork\n\t\t recommendations based on a hybrid combination of artist\n\t\t preference, curated attributes, deep neural visual features\n\t\t and manually-engineered visual features produce the best\n\t\t performance. Moreover, we discuss the trade-off between\n\t\t automatically obtained DNN features and manually-engineered\n\t\t visual features for the purpose of explainability, as well\n\t\t as the impact of user profile size on predictions. Our\n\t\t research informs the development of next-generation\n\t\t content-based artwork recommenders which rely on different\n\t\t types of data, from text to multimedia.},\n url\t\t= {https://link.springer.com/article/10.1007/s11257-018-9206-9}\n}\n\n","author_short":["Messina, P.","Dominguez, V.","Parra, D.","Trattner, C.","Soto, A."],"key":"messina:etal:2019","id":"messina:etal:2019","bibbaseid":"messina-dominguez-parra-trattner-soto-contentbasedartworkrecommendationintegratingpaintingmetadatawithneuralandmanuallyengineeredvisualfeatures-2019","role":"author","urls":{"Paper":"https://link.springer.com/article/10.1007/s11257-018-9206-9"},"metadata":{"authorlinks":{"soto, a":"https://asoto.ing.puc.cl/publications/"}},"downloads":1},"bibtype":"article","biburl":"https://raw.githubusercontent.com/ialab-puc/ialab.ing.puc.cl/master/pubs.bib","creationDate":"2019-06-25T19:39:31.748Z","downloads":1,"keywords":[],"search_terms":["content","based","artwork","recommendation","integrating","painting","metadata","neural","manually","engineered","visual","features","messina","dominguez","parra","trattner","soto"],"title":"Content-based artwork recommendation: integrating painting metadata with neural and manually-engineered visual features","year":2019,"dataSources":["3YPRCmmijLqF4qHXd","sg6yZ29Z2xB5xP79R","m8qFBfFbjk9qWjcmJ","QjT2DEZoWmQYxjHXS"]}