{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/the-sample-complexity-of-online-one-class","title":"The Sample Complexity of Online One-Class Collaborative Filtering","arxiv_id":"1706.00061","date":"2017-05-31","proceeding":"ICML 2017 8","authors":["Reinhard Heckel","Kannan Ramchandran"],"abstract":"We consider the online one-class collaborative filtering (CF) problem that\nconsists of recommending items to users over time in an online fashion based on\npositive ratings only. This problem arises when users respond only occasionally\nto a recommendation with a positive rating, and never with a negative one. We\nstudy the impact of the probability of a user responding to a recommendation,\np_f, on the sample complexity, i.e., the number of ratings required to make\n`good' recommendations, and ask whether receiving positive and negative\nratings, instead of positive ratings only, improves the sample complexity. Both\nquestions arise in the design of recommender systems. We introduce a simple\nprobabilistic user model, and analyze the performance of an online user-based\nCF algorithm. We prove that after an initial cold start phase, where\nrecommendations are invested in exploring the user's preferences, this\nalgorithm makes---up to a fraction of the recommendations required for updating\nthe user's preferences---perfect recommendations. The number of ratings\nrequired for the cold start phase is nearly proportional to 1/p_f, and that for\nupdating the user's preferences is essentially independent of p_f. As a\nconsequence we find that, receiving positive and negative ratings instead of\nonly positive ones improves the number of ratings required for initial\nexploration by a factor of 1/p_f, which can be significant.","url_abs":"http://arxiv.org/abs/1706.00061v1","url_pdf":"http://arxiv.org/pdf/1706.00061v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"the-sample-complexity-of-online-one-class","repo_url":"https://github.com/Atomu2014/product-nets-distributed","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"collaborative-filtering","task_name":"Collaborative Filtering"},{"task_slug":"recommendation-systems","task_name":"Recommendation Systems"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1706.00061","atlas_url":"https://app.syntology.ai/?focus=1706.00061","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}