Improving Existing Collaborative Filtering Recommendations via Serendipity-Based Algorithm. Yang, Y, Xu, Y, Wang, E, Han, J, & Yu, Z IEEE Trans. Multimedia, 20(7):1888–1900, July, 2018. ![Improving Existing Collaborative Filtering Recommendations via Serendipity-Based Algorithm [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex In this paper, we study how to address the sparsity, accuracy and serendipity issues of top-N recommendation with collaborative filtering (CF). Existing studies commonly use rated items (which form only a small section in a rating matrix) or import some additional information (e.g., details about the items and users) to improve the performance of CF. Unlike these methods, we propose a novel notion towards a huge amount of unrated items: serendipity item. By utilizing serendipity items, we propose concise satisfaction and interest injection (CSII), a method that can effectively find interesting, satisfying, and serendipitous items in unrated items. By preventing uninteresting and unsatisfying items to be recommended as top-N items, this concise-but-novel method improves accuracy and recommendation quality (especially serendipity) substantially. Meanwhile, it can address the sparsity and cold-start issues by enriching the rating matrix in CF without additional information. As our method tackles rating matrix before recommendation procedure, it can be applied to most existing CF methods, such as item-based CF, user-based CF and matrix factorization-based CF. Through comprehensive experiments using abundant real-world datasets with LensKit implementation, we successfully demonstrate that our solution improves the performance of existing CF methods consistently and universally. Moreover, comparing with baseline methods, CSII can extract uninteresting items more carefully and cautiously, avoiding potential items inferred by mistake.
@article{yang_improving_2018,
title = {Improving {Existing} {Collaborative} {Filtering} {Recommendations} via {Serendipity}-{Based} {Algorithm}},
volume = {20},
issn = {1520-9210},
url = {http://dx.doi.org/10.1109/TMM.2017.2779043},
doi = {10.1109/TMM.2017.2779043},
abstract = {In this paper, we study how to address the sparsity, accuracy and
serendipity issues of top-N recommendation with collaborative filtering
(CF). Existing studies commonly use rated items (which form only a small
section in a rating matrix) or import some additional information (e.g.,
details about the items and users) to improve the performance of CF.
Unlike these methods, we propose a novel notion towards a huge amount of
unrated items: serendipity item. By utilizing serendipity items, we
propose concise satisfaction and interest injection (CSII), a method that
can effectively find interesting, satisfying, and serendipitous items in
unrated items. By preventing uninteresting and unsatisfying items to be
recommended as top-N items, this concise-but-novel method improves
accuracy and recommendation quality (especially serendipity)
substantially. Meanwhile, it can address the sparsity and cold-start
issues by enriching the rating matrix in CF without additional
information. As our method tackles rating matrix before recommendation
procedure, it can be applied to most existing CF methods, such as
item-based CF, user-based CF and matrix factorization-based CF. Through
comprehensive experiments using abundant real-world datasets with LensKit
implementation, we successfully demonstrate that our solution improves the
performance of existing CF methods consistently and universally. Moreover,
comparing with baseline methods, CSII can extract uninteresting items more
carefully and cautiously, avoiding potential items inferred by mistake.},
number = {7},
journal = {IEEE Trans. Multimedia},
author = {Yang, Y and Xu, Y and Wang, E and Han, J and Yu, Z},
month = jul,
year = {2018},
keywords = {CF methods, CSII, Collaboration, Collaborative filtering, Computer science, Data mining, Lifting equipment, Multimedia communication, Recommender systems, cold-start issues, collaborative filtering, collaborative filtering recommendations, concise satisfaction and interest injection, item-based CF, matrix decomposition, matrix factorization, matrix factorization-based CF, rating matrix, recommendation quality, recommender systems, serendipitous recommendation, serendipity item, top-N items, top-N recommendation, unrated items, user-based CF},
pages = {1888--1900},
}
Downloads: 0
{"_id":"p3b9TgDDWBvEHiQfu","bibbaseid":"yang-xu-wang-han-yu-improvingexistingcollaborativefilteringrecommendationsviaserendipitybasedalgorithm-2018","authorIDs":[],"author_short":["Yang, Y","Xu, Y","Wang, E","Han, J","Yu, Z"],"bibdata":{"bibtype":"article","type":"article","title":"Improving Existing Collaborative Filtering Recommendations via Serendipity-Based Algorithm","volume":"20","issn":"1520-9210","url":"http://dx.doi.org/10.1109/TMM.2017.2779043","doi":"10.1109/TMM.2017.2779043","abstract":"In this paper, we study how to address the sparsity, accuracy and serendipity issues of top-N recommendation with collaborative filtering (CF). Existing studies commonly use rated items (which form only a small section in a rating matrix) or import some additional information (e.g., details about the items and users) to improve the performance of CF. Unlike these methods, we propose a novel notion towards a huge amount of unrated items: serendipity item. By utilizing serendipity items, we propose concise satisfaction and interest injection (CSII), a method that can effectively find interesting, satisfying, and serendipitous items in unrated items. By preventing uninteresting and unsatisfying items to be recommended as top-N items, this concise-but-novel method improves accuracy and recommendation quality (especially serendipity) substantially. Meanwhile, it can address the sparsity and cold-start issues by enriching the rating matrix in CF without additional information. As our method tackles rating matrix before recommendation procedure, it can be applied to most existing CF methods, such as item-based CF, user-based CF and matrix factorization-based CF. Through comprehensive experiments using abundant real-world datasets with LensKit implementation, we successfully demonstrate that our solution improves the performance of existing CF methods consistently and universally. Moreover, comparing with baseline methods, CSII can extract uninteresting items more carefully and cautiously, avoiding potential items inferred by mistake.","number":"7","journal":"IEEE Trans. Multimedia","author":[{"propositions":[],"lastnames":["Yang"],"firstnames":["Y"],"suffixes":[]},{"propositions":[],"lastnames":["Xu"],"firstnames":["Y"],"suffixes":[]},{"propositions":[],"lastnames":["Wang"],"firstnames":["E"],"suffixes":[]},{"propositions":[],"lastnames":["Han"],"firstnames":["J"],"suffixes":[]},{"propositions":[],"lastnames":["Yu"],"firstnames":["Z"],"suffixes":[]}],"month":"July","year":"2018","keywords":"CF methods, CSII, Collaboration, Collaborative filtering, Computer science, Data mining, Lifting equipment, Multimedia communication, Recommender systems, cold-start issues, collaborative filtering, collaborative filtering recommendations, concise satisfaction and interest injection, item-based CF, matrix decomposition, matrix factorization, matrix factorization-based CF, rating matrix, recommendation quality, recommender systems, serendipitous recommendation, serendipity item, top-N items, top-N recommendation, unrated items, user-based CF","pages":"1888–1900","bibtex":"@article{yang_improving_2018,\n\ttitle = {Improving {Existing} {Collaborative} {Filtering} {Recommendations} via {Serendipity}-{Based} {Algorithm}},\n\tvolume = {20},\n\tissn = {1520-9210},\n\turl = {http://dx.doi.org/10.1109/TMM.2017.2779043},\n\tdoi = {10.1109/TMM.2017.2779043},\n\tabstract = {In this paper, we study how to address the sparsity, accuracy and\nserendipity issues of top-N recommendation with collaborative filtering\n(CF). Existing studies commonly use rated items (which form only a small\nsection in a rating matrix) or import some additional information (e.g.,\ndetails about the items and users) to improve the performance of CF.\nUnlike these methods, we propose a novel notion towards a huge amount of\nunrated items: serendipity item. By utilizing serendipity items, we\npropose concise satisfaction and interest injection (CSII), a method that\ncan effectively find interesting, satisfying, and serendipitous items in\nunrated items. By preventing uninteresting and unsatisfying items to be\nrecommended as top-N items, this concise-but-novel method improves\naccuracy and recommendation quality (especially serendipity)\nsubstantially. Meanwhile, it can address the sparsity and cold-start\nissues by enriching the rating matrix in CF without additional\ninformation. As our method tackles rating matrix before recommendation\nprocedure, it can be applied to most existing CF methods, such as\nitem-based CF, user-based CF and matrix factorization-based CF. Through\ncomprehensive experiments using abundant real-world datasets with LensKit\nimplementation, we successfully demonstrate that our solution improves the\nperformance of existing CF methods consistently and universally. Moreover,\ncomparing with baseline methods, CSII can extract uninteresting items more\ncarefully and cautiously, avoiding potential items inferred by mistake.},\n\tnumber = {7},\n\tjournal = {IEEE Trans. Multimedia},\n\tauthor = {Yang, Y and Xu, Y and Wang, E and Han, J and Yu, Z},\n\tmonth = jul,\n\tyear = {2018},\n\tkeywords = {CF methods, CSII, Collaboration, Collaborative filtering, Computer science, Data mining, Lifting equipment, Multimedia communication, Recommender systems, cold-start issues, collaborative filtering, collaborative filtering recommendations, concise satisfaction and interest injection, item-based CF, matrix decomposition, matrix factorization, matrix factorization-based CF, rating matrix, recommendation quality, recommender systems, serendipitous recommendation, serendipity item, top-N items, top-N recommendation, unrated items, user-based CF},\n\tpages = {1888--1900},\n}\n\n","author_short":["Yang, Y","Xu, Y","Wang, E","Han, J","Yu, Z"],"key":"yang_improving_2018","id":"yang_improving_2018","bibbaseid":"yang-xu-wang-han-yu-improvingexistingcollaborativefilteringrecommendationsviaserendipitybasedalgorithm-2018","role":"author","urls":{"Paper":"http://dx.doi.org/10.1109/TMM.2017.2779043"},"keyword":["CF methods","CSII","Collaboration","Collaborative filtering","Computer science","Data mining","Lifting equipment","Multimedia communication","Recommender systems","cold-start issues","collaborative filtering","collaborative filtering recommendations","concise satisfaction and interest injection","item-based CF","matrix decomposition","matrix factorization","matrix factorization-based CF","rating matrix","recommendation quality","recommender systems","serendipitous recommendation","serendipity item","top-N items","top-N recommendation","unrated items","user-based CF"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"article","biburl":"https://api.zotero.org/users/6655/collections/TJPPJ92X/items?key=VFvZhZXIoHNBbzoLZ1IM2zgf&format=bibtex&limit=100","creationDate":"2020-03-27T02:34:35.374Z","downloads":0,"keywords":["cf methods","csii","collaboration","collaborative filtering","computer science","data mining","lifting equipment","multimedia communication","recommender systems","cold-start issues","collaborative filtering","collaborative filtering recommendations","concise satisfaction and interest injection","item-based cf","matrix decomposition","matrix factorization","matrix factorization-based cf","rating matrix","recommendation quality","recommender systems","serendipitous recommendation","serendipity item","top-n items","top-n recommendation","unrated items","user-based cf"],"search_terms":["improving","existing","collaborative","filtering","recommendations","via","serendipity","based","algorithm","yang","xu","wang","han","yu"],"title":"Improving Existing Collaborative Filtering Recommendations via Serendipity-Based Algorithm","year":2018,"dataSources":["5Dp4QphkvpvNA33zi","jfoasiDDpStqkkoZB","BiuuFc45aHCgJqDLY"]}