Recommender response to diversity and popularity bias in user profiles. Channamsetty, S. & Ekstrand, M. D In Proceedings of the 30th Florida artificial intelligence research society conference, May, 2017. AAAI Press. ![Recommender response to diversity and popularity bias in user profiles [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper abstract bibtex 1 download Recommender system evaluation usually focuses on the overall effectiveness of the algorithms, either in terms of measurable accuracy or ability to deliver user satisfaction or improve business metrics. When additional factors are considered, such as the diversity or novelty of the recommendations, the focus typically remains on the algorithm’s overall performance. We examine the relationship of the recommender’s output characteristics – accuracy, popularity (as an inverse of novelty), and diversity – to characteristics of the user’s rating profile. The aims of this analysis are twofold: (1) to probe the conditions under which common algorithms produce more or less diverse or popular recommendations, and (2) to determine if these personalized recommender algorithms reflect a user’s preference for diversity or novelty. We trained recommenders on the MovieLens data and looked for correlation between the user profile and the recommender’s output for both diversity and popularity bias using different metrics. We find that the diversity and popularity of movies in users’ profiles has little impact on the recommendations they receive.
@inproceedings{channamsetty_recommender_2017,
title = {Recommender response to diversity and popularity bias in user profiles},
url = {https://aaai.org/papers/657-flairs-2017-15524/},
abstract = {Recommender system evaluation usually focuses on the overall effectiveness
of the algorithms, either in terms of measurable accuracy or ability to
deliver user satisfaction or improve business metrics. When additional
factors are considered, such as the diversity or novelty of the
recommendations, the focus typically remains on the algorithm’s overall
performance. We examine the relationship of the recommender’s output
characteristics – accuracy, popularity (as an inverse of novelty), and
diversity – to characteristics of the user’s rating profile. The aims of
this analysis are twofold: (1) to probe the conditions under which common
algorithms produce more or less diverse or popular recommendations, and
(2) to determine if these personalized recommender algorithms reflect a
user’s preference for diversity or novelty. We trained recommenders on the
MovieLens data and looked for correlation between the user profile and the
recommender’s output for both diversity and popularity bias using
different metrics. We find that the diversity and popularity of movies in
users’ profiles has little impact on the recommendations they receive.},
urldate = {2017-05-29},
booktitle = {Proceedings of the 30th {Florida} artificial intelligence research society conference},
publisher = {AAAI Press},
author = {Channamsetty, Sushma and Ekstrand, Michael D},
month = may,
year = {2017},
}
Downloads: 1
{"_id":"dozZFJivQpQnnkCdZ","bibbaseid":"channamsetty-ekstrand-recommenderresponsetodiversityandpopularitybiasinuserprofiles-2017","authorIDs":[],"author_short":["Channamsetty, S.","Ekstrand, M. D"],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Recommender response to diversity and popularity bias in user profiles","url":"https://aaai.org/papers/657-flairs-2017-15524/","abstract":"Recommender system evaluation usually focuses on the overall effectiveness of the algorithms, either in terms of measurable accuracy or ability to deliver user satisfaction or improve business metrics. When additional factors are considered, such as the diversity or novelty of the recommendations, the focus typically remains on the algorithm’s overall performance. We examine the relationship of the recommender’s output characteristics – accuracy, popularity (as an inverse of novelty), and diversity – to characteristics of the user’s rating profile. The aims of this analysis are twofold: (1) to probe the conditions under which common algorithms produce more or less diverse or popular recommendations, and (2) to determine if these personalized recommender algorithms reflect a user’s preference for diversity or novelty. We trained recommenders on the MovieLens data and looked for correlation between the user profile and the recommender’s output for both diversity and popularity bias using different metrics. We find that the diversity and popularity of movies in users’ profiles has little impact on the recommendations they receive.","urldate":"2017-05-29","booktitle":"Proceedings of the 30th Florida artificial intelligence research society conference","publisher":"AAAI Press","author":[{"propositions":[],"lastnames":["Channamsetty"],"firstnames":["Sushma"],"suffixes":[]},{"propositions":[],"lastnames":["Ekstrand"],"firstnames":["Michael","D"],"suffixes":[]}],"month":"May","year":"2017","bibtex":"@inproceedings{channamsetty_recommender_2017,\n\ttitle = {Recommender response to diversity and popularity bias in user profiles},\n\turl = {https://aaai.org/papers/657-flairs-2017-15524/},\n\tabstract = {Recommender system evaluation usually focuses on the overall effectiveness\nof the algorithms, either in terms of measurable accuracy or ability to\ndeliver user satisfaction or improve business metrics. When additional\nfactors are considered, such as the diversity or novelty of the\nrecommendations, the focus typically remains on the algorithm’s overall\nperformance. We examine the relationship of the recommender’s output\ncharacteristics – accuracy, popularity (as an inverse of novelty), and\ndiversity – to characteristics of the user’s rating profile. The aims of\nthis analysis are twofold: (1) to probe the conditions under which common\nalgorithms produce more or less diverse or popular recommendations, and\n(2) to determine if these personalized recommender algorithms reflect a\nuser’s preference for diversity or novelty. We trained recommenders on the\nMovieLens data and looked for correlation between the user profile and the\nrecommender’s output for both diversity and popularity bias using\ndifferent metrics. We find that the diversity and popularity of movies in\nusers’ profiles has little impact on the recommendations they receive.},\n\turldate = {2017-05-29},\n\tbooktitle = {Proceedings of the 30th {Florida} artificial intelligence research society conference},\n\tpublisher = {AAAI Press},\n\tauthor = {Channamsetty, Sushma and Ekstrand, Michael D},\n\tmonth = may,\n\tyear = {2017},\n}\n\n","author_short":["Channamsetty, S.","Ekstrand, M. D"],"key":"channamsetty_recommender_2017","id":"channamsetty_recommender_2017","bibbaseid":"channamsetty-ekstrand-recommenderresponsetodiversityandpopularitybiasinuserprofiles-2017","role":"author","urls":{"Paper":"https://aaai.org/papers/657-flairs-2017-15524/"},"metadata":{"authorlinks":{}},"downloads":1},"bibtype":"inproceedings","biburl":"https://api.zotero.org/users/6655/collections/TJPPJ92X/items?key=VFvZhZXIoHNBbzoLZ1IM2zgf&format=bibtex&limit=100","creationDate":"2020-03-27T02:34:35.366Z","downloads":1,"keywords":[],"search_terms":["recommender","response","diversity","popularity","bias","user","profiles","channamsetty","ekstrand"],"title":"Recommender response to diversity and popularity bias in user profiles","year":2017,"dataSources":["5Dp4QphkvpvNA33zi","jfoasiDDpStqkkoZB","BiuuFc45aHCgJqDLY"]}