{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,8]],"date-time":"2026-03-08T00:55:20Z","timestamp":1772931320444,"version":"3.50.1"},"reference-count":66,"publisher":"Association for Computing Machinery (ACM)","issue":"6","license":[{"start":{"date-parts":[[2024,6,25]],"date-time":"2024-06-25T00:00:00Z","timestamp":1719273600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["62302345 and U23A20305"],"award-info":[{"award-number":["62302345 and U23A20305"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]},{"name":"Natural Science Foundation of Hubei Province, China","award":["2023AFB192 and 2023BAB160"],"award-info":[{"award-number":["2023AFB192 and 2023BAB160"]}]}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Inf. Syst."],"published-print":{"date-parts":[[2024,11,30]]},"abstract":"<jats:p>User behavior data, such as ratings and clicks, has been widely used to build personalizing models for recommender systems. However, many unflattering factors\u00a0(e.g., popularity, ranking position, users\u2019 selection) significantly affect the performance of the learned recommendation model. Most existing work on unbiased recommendation addressed these biases from sample granularity\u00a0(e.g., sample reweighting, data augmentation) or from the perspective of representation learning\u00a0(e.g., bias-modeling). However, these methods are usually designed for a specific bias, lacking the universal capability to handle complex situations where multiple biases co-exist. Besides, rare work frees itself from laborious and sophisticated debiasing configurations\u00a0(e.g., propensity scores, imputed values, or user behavior-generating process).<\/jats:p>\n          <jats:p>\n            Towards this research gap, in this article, we propose a universal\n            <jats:bold>G<\/jats:bold>\n            enerative framework for\n            <jats:bold>B<\/jats:bold>\n            ias\n            <jats:bold>D<\/jats:bold>\n            isentanglement termed as\n            <jats:bold>GBD<\/jats:bold>\n            , constantly generating calibration perturbations for the intermediate representations during training to keep them from being affected by the bias. Specifically, a bias-identifier that tries to retrieve the bias-related information from the representations is first introduced. Subsequently, the calibration perturbations are generated to significantly deteriorate the bias-identifier\u2019s performance, making the bias gradually disentangled from the calibrated representations. Therefore, without relying on notorious debiasing configurations, a bias-agnostic model is obtained under the guidance of the bias identifier. We further present its universality by subsuming the representative biases and their mixture under the proposed framework. Finally, extensive experiments on the real-world, synthetic, and semi-synthetic datasets have demonstrated the superiority of the proposed approach against a wide range of recommendation debiasing methods. The code is available at\n            <jats:ext-link xmlns:xlink=\"http:\/\/www.w3.org\/1999\/xlink\" ext-link-type=\"uri\" xlink:href=\"https:\/\/github.com\/Zhidan-Wang\/GBD\">https:\/\/github.com\/Zhidan-Wang\/GBD<\/jats:ext-link>\n            .\n          <\/jats:p>","DOI":"10.1145\/3655617","type":"journal-article","created":{"date-parts":[[2024,4,2]],"date-time":"2024-04-02T13:39:19Z","timestamp":1712065159000},"page":"1-30","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":6,"title":["Toward Bias-Agnostic Recommender Systems: A Universal Generative Framework"],"prefix":"10.1145","volume":"42","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9550-5199","authenticated-orcid":false,"given":"Zhidan","family":"Wang","sequence":"first","affiliation":[{"name":"Department of computer science and technology, Tsinghua University, Beijing, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6755-871X","authenticated-orcid":false,"given":"Lixin","family":"Zou","sequence":"additional","affiliation":[{"name":"School of Cyber Science and Engineering, Wuhan University, Wuhan, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3144-6374","authenticated-orcid":false,"given":"Chenliang","family":"Li","sequence":"additional","affiliation":[{"name":"School of Cyber Science and Engineering, Wuhan University, Wuhan, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9212-1947","authenticated-orcid":false,"given":"Shuaiqiang","family":"Wang","sequence":"additional","affiliation":[{"name":"Baidu Inc, Beijing China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0144-1775","authenticated-orcid":false,"given":"Xu","family":"Chen","sequence":"additional","affiliation":[{"name":"Beijing Key Laboratory, Big Data Management and Analysis Methods, Beijing, China and Gaoling School of Artificial Intelligence, Renmin University of China, Beijing China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0684-6205","authenticated-orcid":false,"given":"Dawei","family":"Yin","sequence":"additional","affiliation":[{"name":"Baidu Inc, Beijing China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1260-4982","authenticated-orcid":false,"given":"Weidong","family":"Liu","sequence":"additional","affiliation":[{"name":"Dept. of Computer Science &amp; Technology, Tsinghua University, Beijing, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"320","published-online":{"date-parts":[[2024,6,25]]},"reference":[{"key":"e_1_3_3_2_2","doi-asserted-by":"publisher","DOI":"10.1145\/3109859.3109912"},{"key":"e_1_3_3_3_2","volume-title":"Proceedings of the 32nd International Flairs Conference","author":"Abdollahpouri Himan","year":"2019","unstructured":"Himan Abdollahpouri, Robin Burke, and Bamshad Mobasher. 2019. Managing popularity bias in recommender systems with personalized re-ranking. In Proceedings of the 32nd International Flairs Conference."},{"key":"e_1_3_3_4_2","article-title":"Multi-sided exposure bias in recommendation","author":"Abdollahpouri Himan","year":"2020","unstructured":"Himan Abdollahpouri and Masoud Mansoury. 2020. Multi-sided exposure bias in recommendation. arXiv:2006.15772. Retrieved from https:\/\/arxiv.org\/abs\/2006.15772","journal-title":"arXiv:2006.15772"},{"key":"e_1_3_3_5_2","doi-asserted-by":"publisher","DOI":"10.1145\/3331184.3331202"},{"key":"e_1_3_3_6_2","doi-asserted-by":"publisher","DOI":"10.1145\/3308558.3313697"},{"key":"e_1_3_3_7_2","doi-asserted-by":"publisher","DOI":"10.1145\/3209978.3209986"},{"key":"e_1_3_3_8_2","article-title":"Domain-adversarial neural networks","author":"Ajakan Hana","year":"2014","unstructured":"Hana Ajakan, Pascal Germain, Hugo Larochelle, Fran\u00e7ois Laviolette, and Mario Marchand. 2014. Domain-adversarial neural networks. arXiv:1412.4446. Retrieved from https:\/\/arxiv.org\/abs\/1412.4446","journal-title":"arXiv:1412.4446"},{"key":"e_1_3_3_9_2","doi-asserted-by":"publisher","DOI":"10.1080\/00401706.1971.10488811"},{"key":"e_1_3_3_10_2","article-title":"Multi-step adversarial perturbations on recommender systems embeddings","author":"Anelli Vito Walter","year":"2020","unstructured":"Vito Walter Anelli, Alejandro Bellog\u00edn, Yashar Deldjoo, Tommaso Di Noia, and Felice Antonio Merra. 2020. Multi-step adversarial perturbations on recommender systems embeddings. arXiv:2010.01329. Retrieved from https:\/\/arxiv.org\/abs\/2010.01329","journal-title":"arXiv:2010.01329"},{"key":"e_1_3_3_11_2","doi-asserted-by":"publisher","DOI":"10.1145\/3336191.3371832"},{"key":"e_1_3_3_12_2","doi-asserted-by":"publisher","DOI":"10.1007\/s10994-009-5152-4"},{"key":"e_1_3_3_13_2","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2017.49"},{"key":"e_1_3_3_14_2","doi-asserted-by":"publisher","DOI":"10.1145\/3298689.3346987"},{"key":"e_1_3_3_15_2","article-title":"AutoDebias: Learning to debias for recommendation","author":"Chen Jiawei","year":"2021","unstructured":"Jiawei Chen, Hande Dong, Yang Qiu, Xiangnan He, Xin Xin, Liang Chen, Guli Lin, and Keping Yang. 2021. AutoDebias: Learning to debias for recommendation. arXiv:2105.04170. Retrieved from https:\/\/arxiv.org\/abs\/2105.04170","journal-title":"arXiv:2105.04170"},{"key":"e_1_3_3_16_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICDM.2018.00018"},{"key":"e_1_3_3_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/3077136.3080797"},{"key":"e_1_3_3_18_2","doi-asserted-by":"publisher","DOI":"10.1145\/3477495.3531952"},{"key":"e_1_3_3_19_2","doi-asserted-by":"publisher","DOI":"10.1145\/3397271.3401043"},{"key":"e_1_3_3_20_2","doi-asserted-by":"publisher","DOI":"10.1145\/3437963.3441757"},{"key":"e_1_3_3_21_2","doi-asserted-by":"publisher","DOI":"10.1109\/MSP.2017.2765202"},{"key":"e_1_3_3_22_2","first-page":"1","volume-title":"Proceedings of the 2020 50th Annual IEEE\/IFIP International Conference on Dependable Systems and Networks Workshops","author":"Noia Tommaso Di","year":"2020","unstructured":"Tommaso Di Noia, Daniele Malitesta, and Felice Antonio Merra. 2020. Taamr: Targeted adversarial attack against multimedia recommender systems. In Proceedings of the 2020 50th Annual IEEE\/IFIP International Conference on Dependable Systems and Networks Workshops. IEEE, 1\u20138."},{"key":"e_1_3_3_23_2","doi-asserted-by":"publisher","DOI":"10.1109\/TMM.2018.2887018"},{"key":"e_1_3_3_24_2","article-title":"Censoring representations with an adversary","author":"Edwards Harrison","year":"2015","unstructured":"Harrison Edwards and Amos Storkey. 2015. Censoring representations with an adversary. arXiv:1511.05897. Retrieved from https:\/\/arxiv.org\/abs\/1511.05897","journal-title":"arXiv:1511.05897"},{"key":"e_1_3_3_25_2","first-page":"1180","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Ganin Yaroslav","year":"2015","unstructured":"Yaroslav Ganin and Victor Lempitsky. 2015. Unsupervised domain adaptation by backpropagation. In Proceedings of the International Conference on Machine Learning. PMLR, 1180\u20131189."},{"key":"e_1_3_3_26_2","doi-asserted-by":"publisher","DOI":"10.1145\/2020408.2020425"},{"key":"e_1_3_3_27_2","doi-asserted-by":"publisher","DOI":"10.1145\/3422622"},{"key":"e_1_3_3_28_2","article-title":"Explaining and harnessing adversarial examples","author":"Goodfellow Ian J.","year":"2014","unstructured":"Ian J. Goodfellow, Jonathon Shlens, and Christian Szegedy. 2014. Explaining and harnessing adversarial examples. arXiv:1412.6572. Retrieved from https:\/\/arxiv.org\/abs\/1412.6572","journal-title":"arXiv:1412.6572"},{"key":"e_1_3_3_29_2","doi-asserted-by":"publisher","DOI":"10.1145\/3289600.3291027"},{"key":"e_1_3_3_30_2","doi-asserted-by":"publisher","DOI":"10.1145\/3298689.3347033"},{"key":"e_1_3_3_31_2","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v30i1.9973"},{"key":"e_1_3_3_32_2","doi-asserted-by":"publisher","DOI":"10.1145\/3209978.3209981"},{"key":"e_1_3_3_33_2","doi-asserted-by":"publisher","DOI":"10.1145\/2911451.2911489"},{"key":"e_1_3_3_34_2","first-page":"1512","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Hern\u00e1ndez-Lobato Jos\u00e9 Miguel","year":"2014","unstructured":"Jos\u00e9 Miguel Hern\u00e1ndez-Lobato, Neil Houlsby, and Zoubin Ghahramani. 2014. Probabilistic matrix factorization with non-random missing data. In Proceedings of the International Conference on Machine Learning. PMLR, 1512\u20131520."},{"key":"e_1_3_3_35_2","doi-asserted-by":"publisher","DOI":"10.1145\/3018661.3018699"},{"key":"e_1_3_3_36_2","volume-title":"RecSys Posters","author":"Kamishima Toshihiro","year":"2014","unstructured":"Toshihiro Kamishima, Shotaro Akaho, Hideki Asoh, and Jun Sakuma. 2014. Correcting popularity bias by enhancing recommendation neutrality. RecSys Posters."},{"key":"e_1_3_3_37_2","doi-asserted-by":"publisher","DOI":"10.1109\/MC.2009.263"},{"key":"e_1_3_3_38_2","doi-asserted-by":"publisher","DOI":"10.1201\/9781351251389-8"},{"key":"e_1_3_3_39_2","doi-asserted-by":"publisher","DOI":"10.1145\/3336191.3371841"},{"key":"e_1_3_3_40_2","doi-asserted-by":"publisher","DOI":"10.1145\/3397271.3401083"},{"key":"e_1_3_3_41_2","article-title":"Towards deep learning models resistant to adversarial attacks","author":"Madry Aleksander","year":"2017","unstructured":"Aleksander Madry, Aleksandar Makelov, Ludwig Schmidt, Dimitris Tsipras, and Adrian Vladu. 2017. Towards deep learning models resistant to adversarial attacks. arXiv:1706.06083. Retrieved from https:\/\/arxiv.org\/abs\/1706.06083","journal-title":"arXiv:1706.06083"},{"key":"e_1_3_3_42_2","doi-asserted-by":"publisher","DOI":"10.1145\/3340531.3412152"},{"key":"e_1_3_3_43_2","article-title":"Collaborative filtering and the missing at random assumption","author":"Marlin Benjamin","year":"2012","unstructured":"Benjamin Marlin, Richard S. Zemel, Sam Roweis, and Malcolm Slaney. 2012. Collaborative filtering and the missing at random assumption. arXiv:1206.5267. Retrieved from https:\/\/arxiv.org\/abs\/1206.5267","journal-title":"arXiv:1206.5267"},{"key":"e_1_3_3_44_2","doi-asserted-by":"publisher","DOI":"10.1145\/3569453"},{"key":"e_1_3_3_45_2","doi-asserted-by":"publisher","DOI":"10.1145\/3366423.3380255"},{"key":"e_1_3_3_46_2","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2016.41"},{"key":"e_1_3_3_47_2","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00908"},{"key":"e_1_3_3_48_2","article-title":"BPR: Bayesian personalized ranking from implicit feedback","author":"Rendle Steffen","year":"2012","unstructured":"Steffen Rendle, Christoph Freudenthaler, Zeno Gantner, and Lars Schmidt-Thieme. 2012. BPR: Bayesian personalized ranking from implicit feedback. arXiv:1205.2618. Retrieved from https:\/\/arxiv.org\/abs\/1205.2618","journal-title":"arXiv:1205.2618"},{"key":"e_1_3_3_49_2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW.2016.58"},{"key":"e_1_3_3_50_2","doi-asserted-by":"publisher","DOI":"10.1145\/3397271.3401114"},{"key":"e_1_3_3_51_2","first-page":"1670","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Schnabel Tobias","year":"2016","unstructured":"Tobias Schnabel, Adith Swaminathan, Ashudeep Singh, Navin Chandak, and Thorsten Joachims. 2016. Recommendations as treatments: Debiasing learning and evaluation. In Proceedings of the International Conference on Machine Learning. PMLR, 1670\u20131679."},{"key":"e_1_3_3_52_2","doi-asserted-by":"publisher","DOI":"10.1145\/1835804.1835895"},{"key":"e_1_3_3_53_2","doi-asserted-by":"publisher","DOI":"10.1145\/3308560.3317303"},{"key":"e_1_3_3_54_2","article-title":"Intriguing properties of neural networks","author":"Szegedy Christian","year":"2013","unstructured":"Christian Szegedy, Wojciech Zaremba, Ilya Sutskever, Joan Bruna, Dumitru Erhan, Ian Goodfellow, and Rob Fergus. 2013. Intriguing properties of neural networks. arXiv:1312.6199. Retrieved from https:\/\/arxiv.org\/abs\/1312.6199","journal-title":"arXiv:1312.6199"},{"key":"e_1_3_3_55_2","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2019.2893638"},{"issue":"11","key":"e_1_3_3_56_2","article-title":"Visualizing data using t-SNE.","volume":"9","author":"Maaten Laurens Van der","year":"2008","unstructured":"Laurens Van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-SNE. Journal of Machine Learning Research 9, 11 (2008).","journal-title":"Journal of Machine Learning Research"},{"key":"e_1_3_3_57_2","article-title":"Achieving fairness through adversarial learning: An application to recidivism prediction","author":"Wadsworth Christina","year":"2018","unstructured":"Christina Wadsworth, Francesca Vera, and Chris Piech. 2018. Achieving fairness through adversarial learning: An application to recidivism prediction. arXiv:1807.00199. Retrieved from https:\/\/arxiv.org\/abs\/1807.00199","journal-title":"arXiv:1807.00199"},{"key":"e_1_3_3_58_2","doi-asserted-by":"publisher","DOI":"10.1145\/3331184.3331267"},{"key":"e_1_3_3_59_2","first-page":"6638","volume-title":"Proceedings of the International Conference on Machine Learning","author":"Wang Xiaojie","year":"2019","unstructured":"Xiaojie Wang, Rui Zhang, Yu Sun, and Jianzhong Qi. 2019. Doubly robust joint learning for recommendation on data missing not at random. In Proceedings of the International Conference on Machine Learning. PMLR, 6638\u20136647."},{"key":"e_1_3_3_60_2","doi-asserted-by":"publisher","DOI":"10.1145\/3447548.3467289"},{"key":"e_1_3_3_61_2","doi-asserted-by":"publisher","DOI":"10.3354\/cr030079"},{"key":"e_1_3_3_62_2","doi-asserted-by":"publisher","DOI":"10.1109\/IJCNN.2019.8851902"},{"key":"e_1_3_3_63_2","article-title":"Causal intervention for leveraging popularity bias in recommendation","author":"Zhang Yang","year":"2021","unstructured":"Yang Zhang, Fuli Feng, Xiangnan He, Tianxin Wei, Chonggang Song, Guohui Ling, and Yongdong Zhang. 2021. Causal intervention for leveraging popularity bias in recommendation. arXiv:2105.06067. Retrieved from https:\/\/arxiv.org\/abs\/2105.06067","journal-title":"arXiv:2105.06067"},{"key":"e_1_3_3_64_2","article-title":"Popularity bias is not always evil: Disentangling benign and harmful bias for recommendation","author":"Zhao Zihao","year":"2022","unstructured":"Zihao Zhao, Jiawei Chen, Sheng Zhou, Xiangnan He, Xuezhi Cao, Fuzheng Zhang, and Wei Wu. 2022. Popularity bias is not always evil: Disentangling benign and harmful bias for recommendation. IEEE Transactions on Knowledge and Data Engineering (2022).","journal-title":"IEEE Transactions on Knowledge and Data Engineering"},{"key":"e_1_3_3_65_2","article-title":"Disentangling user interest and popularity bias for recommendation with causal embedding","author":"Zheng Yu","year":"2020","unstructured":"Yu Zheng, Chen Gao, Xiang Li, Xiangnan He, Yong Li, and Depeng Jin. 2020. Disentangling user interest and popularity bias for recommendation with causal embedding. arXiv:2006.11011. Retrieved from https:\/\/arxiv.org\/abs\/2006.11011","journal-title":"arXiv:2006.11011"},{"key":"e_1_3_3_66_2","doi-asserted-by":"publisher","DOI":"10.1145\/3437963.3441820"},{"key":"e_1_3_3_67_2","doi-asserted-by":"publisher","DOI":"10.1145\/3442381.3450098"}],"container-title":["ACM Transactions on Information Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3655617","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3655617","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,19]],"date-time":"2025-06-19T00:03:46Z","timestamp":1750291426000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3655617"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,6,25]]},"references-count":66,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2024,11,30]]}},"alternative-id":["10.1145\/3655617"],"URL":"https:\/\/doi.org\/10.1145\/3655617","relation":{},"ISSN":["1046-8188","1558-2868"],"issn-type":[{"value":"1046-8188","type":"print"},{"value":"1558-2868","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,6,25]]},"assertion":[{"value":"2023-05-04","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-03-17","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2024-06-25","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}