{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,12]],"date-time":"2026-03-12T22:57:23Z","timestamp":1773356243391,"version":"3.50.1"},"reference-count":52,"publisher":"Association for Computing Machinery (ACM)","issue":"3","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Inf. Syst."],"published-print":{"date-parts":[[2026,3,31]]},"abstract":"<jats:p>Recommendations usually focus on immediate accuracy metrics like Click-Through Rate (CTR), ignoring user long-term metrics. User retention, which reflects the percentage of today\u2019s users who will return to the system in the next few days, should be paid more attention to. However, most existing methods did not focus on user retention, since their complexity and uncertainty make it extremely hard to discover why a user will or will not return to a system. Recently, a few pioneers have optimized user retention, focusing solely on accuracy without delving into its underlying rationale. This is primarily due to the absence of explicit supervised signals. In this work, we design a Behavior-wise Contrastive Multi-Instance Learning (BCMIL) module, which jointly models clicked and impressed items to capture interpretable user retention. Specifically, we conduct in-depth analyses in real-world scenarios to discover implicit retention-related supervised signals. To model these signals, we design a Forward Supervised Signals Extractor (FSSE) that utilizes a heterogeneous graph, enhancing the reliability of user retention. To mitigate randomness and uncertainty, we propose a Backward Supervised Signals Stabilizer (BSSS) that utilizes overlooked label-part behaviors within each training window to retrospectively guide the training process. Offline and online evaluations of an industrial system verify the effectiveness of our methods.<\/jats:p>","DOI":"10.1145\/3790098","type":"journal-article","created":{"date-parts":[[2026,1,27]],"date-time":"2026-01-27T13:36:14Z","timestamp":1769520974000},"page":"1-35","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Rethinking User Retention Modeling in Recommendation"],"prefix":"10.1145","volume":"44","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8342-7875","authenticated-orcid":false,"given":"Rui","family":"Ding","sequence":"first","affiliation":[{"name":"School of Computer Science and Engineering, Northeastern University, Shenyang, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3170-5647","authenticated-orcid":false,"given":"Ruobing","family":"Xie","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-8793-3315","authenticated-orcid":false,"given":"Xiaobo","family":"Hao","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6184-4771","authenticated-orcid":false,"given":"Xiaochun","family":"Yang","sequence":"additional","affiliation":[{"name":"School of Computer Science and Engineering, Northeastern University, Shenyang, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-0554-0761","authenticated-orcid":false,"given":"Kaikai","family":"Ge","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5685-316X","authenticated-orcid":false,"given":"Xu","family":"Zhang","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-5151-4222","authenticated-orcid":false,"given":"Zhanhui","family":"Kang","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5899-5165","authenticated-orcid":false,"given":"Jie","family":"Zhou","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5471-500X","authenticated-orcid":false,"given":"Leyu","family":"Lin","sequence":"additional","affiliation":[{"name":"Tencent, Beijing, China"}]}],"member":"320","published-online":{"date-parts":[[2026,3,2]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.1145\/3543873.3584640"},{"key":"e_1_3_1_3_2","doi-asserted-by":"publisher","DOI":"10.1145\/3340631.3394856"},{"key":"e_1_3_1_4_2","doi-asserted-by":"publisher","DOI":"10.1145\/2487788.2488090"},{"key":"e_1_3_1_5_2","volume-title":"Recommender Systems Handbook.","author":"Castells Pablo","year":"2021","unstructured":"Pablo Castells, Neil Hurley, and Saul Vargas. 2021. Novelty and diversity in recommender systems. In Recommender Systems Handbook. Francesco Ricci, Lior Rokach, and Bracha Shapira (Eds.), Springer."},{"key":"e_1_3_1_6_2","first-page":"19","volume-title":"Proceedings of the 33rd IEEE International Conference on Data Engineering (ICDE \u201917)","author":"Chen Jun","year":"2017","unstructured":"Jun Chen, Chaokun Wang, Jianmin Wang, and Philip S. Yu. 2017. Recommendation for repeat consumption from user implicit feedback. In Proceedings of the 33rd IEEE International Conference on Data Engineering (ICDE \u201917). IEEE Computer Society, 19\u201320."},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1145\/3308558.3313469"},{"key":"e_1_3_1_8_2","doi-asserted-by":"publisher","DOI":"10.1145\/3460231.3474236"},{"key":"e_1_3_1_9_2","doi-asserted-by":"publisher","DOI":"10.1145\/3038912.3052599"},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1145\/2488388.2488418"},{"key":"e_1_3_1_11_2","doi-asserted-by":"publisher","DOI":"10.1137\/1.9781611976236.10"},{"key":"e_1_3_1_12_2","doi-asserted-by":"publisher","DOI":"10.1145\/3604915.3608818"},{"key":"e_1_3_1_13_2","doi-asserted-by":"publisher","DOI":"10.1145\/2187980.2188207"},{"key":"e_1_3_1_14_2","unstructured":"Ziv Epstein Hause Lin Gordon Pennycook and David G. Rand. 2022. Quantifying attention via dwell time and engagement in a social media browsing environment. arXiv:2209.10464. Retrieved from https:\/\/arxiv.org\/abs\/2209.10464"},{"key":"e_1_3_1_15_2","doi-asserted-by":"publisher","DOI":"10.1145\/3523227.3546767"},{"key":"e_1_3_1_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/3442381.3449848"},{"key":"e_1_3_1_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/3404835.3462976"},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2017\/239"},{"key":"e_1_3_1_19_2","doi-asserted-by":"publisher","DOI":"10.1145\/3295822"},{"key":"e_1_3_1_20_2","unstructured":"Bin Hao Min Zhang Weizhi Ma Shaoyun Shi Xinxing Yu Houzhi Shan Yiqun Liu and Shaoping Ma. 2021. A large-scale rich context query and recommendation dataset in online knowledge-sharing. arXiv:2106.06467. Retrieved from https:\/\/arxiv.org\/abs\/2106.06467"},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/3404835.3463117"},{"key":"e_1_3_1_22_2","doi-asserted-by":"publisher","DOI":"10.1145\/3511808.3557579"},{"key":"e_1_3_1_23_2","doi-asserted-by":"publisher","DOI":"10.1145\/3397271.3401082"},{"key":"e_1_3_1_24_2","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2020\/434"},{"key":"e_1_3_1_25_2","doi-asserted-by":"publisher","DOI":"10.1145\/3308558.3313613"},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1007\/s13740-016-0061-8"},{"key":"e_1_3_1_27_2","doi-asserted-by":"publisher","DOI":"10.1145\/3611659.3615687"},{"key":"e_1_3_1_28_2","doi-asserted-by":"publisher","DOI":"10.1145\/3616855.3635833"},{"key":"e_1_3_1_29_2","first-page":"1","volume-title":"Proceedings of the 7th Joint Workshop on Interfaces and Human Decision Making for Recommender Systems Co-Located with 14th ACM Conference on Recommender Systems (RecSys \u201920)","volume":"2682","author":"Tsai Chun-Hua","year":"2020","unstructured":"Chun-Hua Tsai and Peter Brusilovsky. 2020. User feedback in controllable and explainable social recommender systems: A linguistic analysis. In Proceedings of the 7th Joint Workshop on Interfaces and Human Decision Making for Recommender Systems Co-Located with 14th ACM Conference on Recommender Systems (RecSys \u201920). CEUR Workshop Proceedings, Vol. 2682, CEUR-WS.org, 1\u201313."},{"key":"e_1_3_1_30_2","volume-title":"Proceedings of the 6th International Conference on Learning Representations (ICLR \u201918)","author":"Velickovic Petar","year":"2018","unstructured":"Petar Velickovic, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Li\u00f2, and Yoshua Bengio. 2018. Graph attention networks. In Proceedings of the 6th International Conference on Learning Representations (ICLR \u201918). Conference Track Proceedings."},{"key":"e_1_3_1_31_2","doi-asserted-by":"publisher","DOI":"10.1145\/3077136.3080786"},{"key":"e_1_3_1_32_2","doi-asserted-by":"publisher","DOI":"10.1145\/3627826"},{"key":"e_1_3_1_33_2","doi-asserted-by":"publisher","DOI":"10.5555\/2832415.2832502"},{"key":"e_1_3_1_34_2","doi-asserted-by":"publisher","DOI":"10.1145\/3534678.3539073"},{"key":"e_1_3_1_35_2","doi-asserted-by":"publisher","DOI":"10.1002\/jocb.433"},{"key":"e_1_3_1_36_2","doi-asserted-by":"publisher","DOI":"10.1145\/3485447.3512082"},{"key":"e_1_3_1_37_2","doi-asserted-by":"publisher","DOI":"10.1145\/3132847.3133025"},{"key":"e_1_3_1_38_2","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-00126-0_11"},{"key":"e_1_3_1_39_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477495.3531913"},{"key":"e_1_3_1_40_2","doi-asserted-by":"publisher","DOI":"10.1109\/TBDATA.2021.3103263"},{"key":"e_1_3_1_41_2","doi-asserted-by":"publisher","DOI":"10.1145\/3543873.3584624"},{"issue":"1","key":"e_1_3_1_42_2","first-page":"1","article-title":"Group-based personalized news recommendation with long- and short-term fine-grained matching","volume":"42","author":"Xu Hongyan","year":"2023","unstructured":"Hongyan Xu, Qiyao Peng, Hongtao Liu, Yueheng Sun, and Wenjun Wang. 2023. Group-based personalized news recommendation with long- and short-term fine-grained matching. ACM Transactions on Information Systems 42, 1 (2023), 1\u201327.","journal-title":"ACM Transactions on Information Systems"},{"key":"e_1_3_1_43_2","doi-asserted-by":"publisher","DOI":"10.1145\/3580305.3599473"},{"key":"e_1_3_1_44_2","volume-title":"Proceedings of the 11th International Conference on Learning Representations (ICLR \u201923)","author":"Xue Wanqi","year":"2023","unstructured":"Wanqi Xue, Qingpeng Cai, Ruohan Zhan, Dong Zheng, Peng Jiang, Kun Gai, and Bo An. 2023. ResAct: Reinforcing long-term engagement in sequential recommendation with residual actor. In Proceedings of the 11th International Conference on Learning Representations (ICLR \u201923). OpenReview.net."},{"key":"e_1_3_1_45_2","doi-asserted-by":"publisher","DOI":"10.1145\/3543873.3584628"},{"key":"e_1_3_1_46_2","doi-asserted-by":"publisher","DOI":"10.1145\/2645710.2645724"},{"key":"e_1_3_1_47_2","doi-asserted-by":"publisher","DOI":"10.1145\/3534678.3539040"},{"key":"e_1_3_1_48_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477495.3531829"},{"key":"e_1_3_1_49_2","doi-asserted-by":"publisher","DOI":"10.1145\/3543507.3583418"},{"key":"e_1_3_1_50_2","doi-asserted-by":"publisher","DOI":"10.1145\/3394486.3403329"},{"key":"e_1_3_1_51_2","doi-asserted-by":"publisher","DOI":"10.1145\/3178876.3185994"},{"key":"e_1_3_1_52_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477495.3532030"},{"key":"e_1_3_1_53_2","doi-asserted-by":"publisher","DOI":"10.1145\/3292500.3330668"}],"container-title":["ACM Transactions on Information Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3790098","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,12]],"date-time":"2026-03-12T12:18:32Z","timestamp":1773317912000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3790098"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,2]]},"references-count":52,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2026,3,31]]}},"alternative-id":["10.1145\/3790098"],"URL":"https:\/\/doi.org\/10.1145\/3790098","relation":{},"ISSN":["1046-8188","1558-2868"],"issn-type":[{"value":"1046-8188","type":"print"},{"value":"1558-2868","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,2]]},"assertion":[{"value":"2025-02-09","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-01-11","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2026-03-02","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}