ABSTRACT
Biases and de-biasing in recommender systems (RS) have become a research hotspot recently. This paper reveals an unexplored type of bias, i.e., sentiment bias. Through an empirical study, we find that many RS models provide more accurate recommendations on user/item groups having more positive feedback (i.e., positive users/items) than on user/item groups having more negative feedback (i.e., negative users/items). We show that sentiment bias is different from existing biases such as popularity bias: positive users/items do not have more user feedback (i.e., either more ratings or longer reviews). The existence of sentiment bias leads to low-quality recommendations to critical users and unfair recommendations for niche items. We discuss the factors that cause sentiment bias. Then, to fix the sources of sentiment bias, we propose a general de-biasing framework with three strategies manifesting in different regularizers that can be easily plugged into RS models without changing model architectures. Experiments on various RS models and benchmark datasets have verified the effectiveness of our de-biasing framework. To our best knowledge, sentiment bias and its de-biasing have not been studied before. We hope that this work can help strengthen the study of biases and de-biasing in RS.
Supplemental Material
- Himan Abdollahpouri, Robin Burke, and Bamshad Mobasher. 2017. Controlling Popularity Bias in Learning-to-Rank Recommendation. In RecSys. 42--46.Google Scholar
- Himan Abdollahpouri, Robin Burke, and Bamshad Mobasher. 2019 a. Managing Popularity Bias in Recommender Systems with Personalized Re-Ranking. In FLAIRS Conference. 413--418.Google Scholar
- Himan Abdollahpouri and Masoud Mansoury. 2020. Multi-sided Exposure Bias in Recommendation. In IRS2020@KDD .Google Scholar
- Himan Abdollahpouri, Masoud Mansoury, Robin Burke, and Bamshad Mobasher. 2019 b. The Unfairness of Popularity Bias in Recommendation. In RMSE@RecSys, Vol. 2440.Google Scholar
- Himan Abdollahpouri, Masoud Mansoury, Robin Burke, and Bamshad Mobasher. 2020. The Connection Between Popularity Bias, Calibration, and Fairness in Recommendation. In RecSys. 726--731.Google Scholar
- Charu C. Aggarwal. 2016. Recommender Systems - The Textbook .Springer.Google ScholarDigital Library
- Stephen Bonner and Flavian Vasile. 2018. Causal embeddings for recommendation. In RecSys. 104--112.Google Scholar
- Chong Chen, Min Zhang, Yiqun Liu, and Shaoping Ma. 2018. Neural Attentional Rating Regression with Review-level Explanations. In WWW. 1583--1592.Google Scholar
- Jiawei Chen, Hande Dong, Xiang Wang, Fuli Feng, Meng Wang, and Xiangnan He. 2020 a. Bias and Debias in Recommender System: A Survey and Future Directions. arXiv Preprint (2020). https://arxiv.org/abs/2010.03240Google Scholar
- Zhihong Chen, Rong Xiao, Chenliang Li, Gangfeng Ye, Haochuan Sun, and Hongbo Deng. 2020 b. ESAM: Discriminative Domain Adaptation with Non-Displayed Items to Improve Long-Tail Performance. In SIGIR. 579--588.Google Scholar
- Andrew Collins, Dominika Tkaczyk, Akiko Aizawa, and Jö ran Beel. 2018. A Study of Position Bias in Digital Library Recommender Systems. arXiv Preprint (2018). https://arxiv.org/abs/1802.06565Google Scholar
- Alberto Garc'i a-Durá n, Roberto Gonzalez, Daniel O n oro-Rubio, Mathias Niepert, and Hui Li. 2020. TransRev: Modeling Reviews as Translations from Users to Items. In ECIR, Vol. 12035. 234--248.Google Scholar
- David Godes and Jose C. Silva. 2012. Sequential and Temporal Dynamics of Online Opinion. Marketing Science, Vol. 31, 3 (2012), 448--473.Google ScholarDigital Library
- Huifeng Guo, Jinkai Yu, Qing Liu, Ruiming Tang, and Yuzhou Zhang. 2019. PAL: a position-bias aware learning framework for CTR prediction in live recommender systems. In RecSys. 452--456.Google Scholar
- Ruining He, Wang-Cheng Kang, and Julian J. McAuley. 2017a. Translation-based Recommendation. In RecSys. 161--169.Google Scholar
- Ruining He, Wang-Cheng Kang, and Julian J. McAuley. 2018. Translation-based Recommendation: A Scalable Method for Modeling Sequential Behavior. In IJCAI. 5264--5268.Google Scholar
- Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua. 2017b. Neural Collaborative Filtering. In WWW. 173--182.Google Scholar
- Katja Hofmann, Anne Schuth, Alejandro Bellog'i n, and Maarten de Rijke. 2014. Effects of Position Bias on Click-Based Recommender Evaluation. In ECIR, Vol. 8416. 624--630.Google Scholar
- Dongmin Hyun, Chanyoung Park, Min-Chul Yang, Ilhyeon Song, Jung-Tae Lee, and Hwanjo Yu. 2018. Review Sentiment-Guided Scalable Deep Recommender System. In SIGIR. 965--968.Google Scholar
- Yehuda Koren, Robert M. Bell, and Chris Volinsky. 2009. Matrix Factorization Techniques for Recommender Systems. Computer, Vol. 42, 8 (2009), 30--37.Google ScholarDigital Library
- Dominik Kowald, Markus Schedl, and Elisabeth Lex. 2020. The Unfairness of Popularity Bias in Music Recommendation: A Reproducibility Study. In ECIR, Vol. 12036. 35--42.Google ScholarDigital Library
- Sanjay Krishnan, Jay Patel, Michael J. Franklin, and Ken Goldberg. 2014. A methodology for learning, analyzing, and mitigating social influence bias in recommender systems. In RecSys. 137--144.Google Scholar
- Hui Li, Ye Liu, Nikos Mamoulis, and David S. Rosenblum. 2020. Translation-Based Sequential Recommendation for Complex Users on Sparse Data. IEEE Trans. Knowl. Data Eng., Vol. 32, 8 (2020), 1639--1651.Google ScholarCross Ref
- Dawen Liang, Laurent Charlin, James McInerney, and David M. Blei. 2016. Modeling User Exposure in Recommendation. In WWW. 951--961.Google ScholarDigital Library
- Dugang Liu, Pengxiang Cheng, Zhenhua Dong, Xiuqiang He, Weike Pan, and Zhong Ming. 2020 a. A General Knowledge Distillation Framework for Counterfactual Recommendation via Uniform Data. In SIGIR. 831--840.Google Scholar
- Donghua Liu, Jing Li, Bo Du, Jun Chang, and Rong Gao. 2019. DAML: Dual Attention Mutual Learning between Ratings and Reviews for Item Recommendation. In KDD. 344--352.Google Scholar
- Hongtao Liu, Wenjun Wang, Hongyan Xu, Qiyao Peng, and Pengfei Jiao. 2020 b. Neural Unified Review Recommendation with Cross Attention. In SIGIR. 1789--1792.Google Scholar
- Yiming Liu, Xuezhi Cao, and Yong Yu. 2016. Are You Influenced by Others When Rating?: Improve Rating Prediction by Conformity Modeling. In RecSys. 269--272.Google Scholar
- Hao Ma, Dengyong Zhou, Chao Liu, Michael R. Lyu, and Irwin King. 2011. Recommender systems with social regularization. In WSDM. 287--296.Google Scholar
- Benjamin M. Marlin, Richard S. Zemel, Sam T. Roweis, and Malcolm Slaney. 2007. Collaborative Filtering and the Missing at Random Assumption. In UAI. 267--275.Google Scholar
- Julian J. McAuley, Rahul Pandey, and Jure Leskovec. 2015. Inferring Networks of Substitutable and Complementary Products. In KDD. 785--794.Google Scholar
- Ninareh Mehrabi, Fred Morstatter, Nripsuta Saxena, Kristina Lerman, and Aram Galstyan. 2019. A Survey on Bias and Fairness in Machine Learning. arXiv Preprint (2019). https://arxiv.org/abs/1908.09635Google Scholar
- Rajiv Pasricha and Julian J. McAuley. 2018. Translation-based factorization machines for sequential recommendation. In RecSys. 63--71.Google Scholar
- Steffen Rendle, Walid Krichene, Li Zhang, and John R. Anderson. 2020. Neural Collaborative Filtering vs. Matrix Factorization Revisited. In RecSys. 240--248.Google ScholarDigital Library
- Noveen Sachdeva and Julian J. McAuley. 2020. How Useful are Reviews for Recommendation? A Critical Review and Potential Improvements. In SIGIR. 1845--1848.Google Scholar
- Yuta Saito. 2020. Asymmetric Tri-training for Debiasing Missing-Not-At-Random Explicit Feedback. In SIGIR. 309--318.Google Scholar
- Tobias Schnabel, Adith Swaminathan, Ashudeep Singh, Navin Chandak, and Thorsten Joachims. 2016. Recommendations as Treatments: Debiasing Learning and Evaluation. In ICML, Vol. 48. 1670--1679.Google Scholar
- Harald Steck. 2010. Training and testing of recommender systems on data missing not at random. In KDD. 713--722.Google Scholar
- Tony Sun, Andrew Gaut, Shirlyn Tang, Yuxin Huang, Mai ElSherief, Jieyu Zhao, Diba Mirza, Elizabeth M. Belding, Kai-Wei Chang, and William Yang Wang. 2019. Mitigating Gender Bias in Natural Language Processing: Literature Review. In ACL. 1630--1640.Google Scholar
- Yi Tay, Anh Tuan Luu, and Siu Cheung Hui. 2018. Multi-Pointer Co-Attention Networks for Recommendation. In KDD. 2309--2318.Google ScholarDigital Library
- Laurens van der Maaten and Geoffrey Hinton. 2008. Visualizing Data using t-SNE. J. Mach. Learn. Res., Vol. 9 (2008), 2579--2605.Google Scholar
- Mei Wang and Weihong Deng. 2020. Mitigating Bias in Face Recognition Using Skewness-Aware Reinforcement Learning. In CVPR. 9319--9328.Google Scholar
- Quan Wang, Zhendong Mao, Bin Wang, and Li Guo. 2017. Knowledge Graph Embedding: A Survey of Approaches and Applications. IEEE Trans. Knowl. Data Eng., Vol. 29, 12 (2017), 2724--2743.Google ScholarCross Ref
- Ting Wang and Dashun Wang. 2014. Why Amazon's Ratings Might Mislead You: The Story of Herding Effects. Big Data, Vol. 2, 4 (2014), 196--204.Google ScholarCross Ref
- Yixin Wang, Dawen Liang, Laurent Charlin, and David M. Blei. 2018. The Deconfounded Recommender: A Causal Inference Approach to Recommendation. arXiv Preprint (2018). https://arxiv.org/abs/1808.06581Google Scholar
- Yixin Wang, Dawen Liang, Laurent Charlin, and David M. Blei. 2020. Causal Inference for Recommender Systems. In RecSys. 426--431.Google Scholar
- Tianxin Wei, Fuli Feng, Jiawei Chen, Chufeng Shi, Ziwei Wu, Jinfeng Yi, and Xiangnan He. 2020. Model-Agnostic Counterfactual Reasoning for Eliminating Popularity Bias in Recommender System. arXiv Preprint (2020). https://arxiv.org/abs/2010.15363Google Scholar
- Yin Zhang, Yun He, Jianling Wang, and James Caverlee. 2020. Adaptive Hierarchical Translation-based Sequential Recommendation. In WWW. 2984--2990.Google Scholar
- Hua Zheng, Dong Wang, Qi Zhang, Hang Li, and Tinghao Yang. 2010. Do clicks measure recommendation relevancy?: an empirical user study. In RecSys. 249--252.Google Scholar
- Lei Zheng, Vahid Noroozi, and Philip S. Yu. 2017. Joint Deep Modeling of Users and Items Using Reviews for Recommendation. In WSDM. 425--434.Google Scholar
Index Terms
- Mitigating Sentiment Bias for Recommender Systems
Recommendations
Causal Intervention for Sentiment De-biasing in Recommendation
CIKM '22: Proceedings of the 31st ACM International Conference on Information & Knowledge ManagementBiases and de-biasing in recommender systems have received increasing attention recently. This study focuses on a newly identified bias, i.e., sentiment bias, which is defined as the divergence in recommendation performance between positive users/items ...
Measuring and Mitigating Item Under-Recommendation Bias in Personalized Ranking Systems
SIGIR '20: Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information RetrievalRecommendation algorithms typically build models based on user-item interactions (e.g., clicks, likes, or ratings) to provide a personalized ranked list of items. These interactions are often distributed unevenly over different groups of items due to ...
Acquiring User Information Needs for Recommender Systems
WI-IAT '13: Proceedings of the 2013 IEEE/WIC/ACM International Joint Conferences on Web Intelligence (WI) and Intelligent Agent Technologies (IAT) - Volume 03Most recommender systems attempt to use collaborative filtering, content-based filtering or hybrid approach to recommend items to new users. Collaborative filtering recommends items to new users based on their similar neighbours, and content-based ...
Comments