Debiasing Recommendation with Personal Popularity
Authors: 
Wentao Ning, Reynold Cheng, Xiao Yan, Ben Kao, Nan Huo, Nur Al Hasan Haldar, and Bo TangAuthors Info & Claims
Wentao Ning, Reynold Cheng, Xiao Yan, Ben Kao, Nan Huo, Nur Al Hasan Haldar, and Bo TangAuthors Info & ClaimsMay 2024
Pages 3400 - 3409
Abstract
Global popularity (GP) bias is the phenomenon that popular items are recommended much more frequently than they should be, which goes against the goal of providing personalized recommendations and harms user experience and recommendation accuracy. Many methods have been proposed to reduce GP bias but they fail to notice the fundamental problem of GP, i.e., it considers popularity from a global perspective of all users and uses a single set of popular items, and thus cannot capture the interests of individual users. As such, we propose a user-aware version of item popularity named personal popularity (PP), which identifies different popular items for each user by considering the users that share similar interests. As PP models the preferences of individual users, it naturally helps to produce personalized recommendations and mitigate GP bias. To integrate PP into recommendation, we design a general personal popularity aware counterfactual (PPAC) framework, which adapts easily to existing recommendation models. In particular, PPAC recognizes that PP and GP have both direct and indirect effects on recommendations and controls direct effects with counterfactual inference techniques for unbiased recommendations. All codes and datasets are available at https://github.com/Stevenn9981/PPAC.
Supplemental Material
MP4 File
Video presentation
- Download
- 1599.31 MB
MP4 File
Supplemental video
- Download
- 2.76 MB
References
[1]
H. Abdollahpouri, R. Burke, and B. Mobasher. Controlling popularity bias in learning-to-rank recommendation. In RecSys, pages 42--46, 2017.
[2]
H. Abdollahpouri, R. Burke, and B. Mobasher. Managing popularity bias in recommender systems with personalized re-ranking. In FLAIRS, pages 413--418, 2019.
[3]
H. Abdollahpouri, M. Mansoury, R. Burke, and B. Mobasher. The connection between popularity bias, calibration, and fairness in recommendation. In RecSys, pages 726--731, 2020.
[4]
H. Abdollahpouri, M. Mansoury, R. Burke, B. Mobasher, and E. C. Malthouse. User-centered evaluation of popularity bias in recommender systems. In UMAP, pages 119--129, 2021.
[5]
G. Balloccu, L. Boratto, G. Fenu, and M. Marras. Post processing recommender systems with knowledge graphs for recency, popularity, and diversity of explanations. In SIGIR, pages 646--656, 2022.
[6]
S. Bonner and F. Vasile. Causal embeddings for recommendation. In RecSys, pages 104--112, 2018.
[7]
L. Bottou, J. Peters, J. Q. Candela, D. X. Charles, M. Chickering, E. Portugaly, D. Ray, P. Y. Simard, and E. Snelson. Counterfactual reasoning and learning systems: the example of computational advertising. JMLR, 14(1):3207--3260, 2013.
[8]
R. Cañamares and P. Castells. A probabilistic reformulation of memory-based collaborative filtering: Implications on popularity biases. In SIGIR, pages 215--224, 2017.
[9]
A. J. B. Chaney, B. M. Stewart, and B. E. Engelhardt. How algorithmic confounding in recommendation systems increases homogeneity and decreases utility. In RecSys, pages 224--232, 2018.
[10]
J. Chen, X. Wang, F. Feng, and X. He. Bias issues and solutions in recommender system: Tutorial on the recsys 2021. In RecSys, pages 825--827, 2021.
[11]
L. Chen, H. Zhang, J. Xiao, X. He, S. Pu, and S. Chang. Counterfactual critic multi-agent training for scene graph generation. In ICCV, pages 4612--4622, 2019.
[12]
Z. Chen, J. Wu, C. Li, J. Chen, R. Xiao, and B. Zhao. Co-training disentangled domain adaptation network for leveraging popularity bias in recommenders. In SIGIR, pages 60--69, 2022.
[13]
Z. Chen, R. Xiao, C. Li, G. Ye, H. Sun, and H. Deng. ESAM: discriminative domain adaptation with non-displayed items to improve long-tail performance. In SIGIR, pages 579--588, 2020.
[14]
F. Feng, W. Huang, X. He, X. Xin, Q. Wang, and T. Chua. Should graph convolution trust neighbors? A simple causal inference method. In SIGIR, pages 1208--1218, 2021.
[15]
F. Feng, J. Zhang, X. He, H. Zhang, and T. Chua. Empowering language understanding with counterfactual reasoning. In ACL/IJCNLP, volume ACL/IJCNLP 2021 of Findings of ACL, pages 2226--2236, 2021.
[16]
Z. Fu, Y. Xian, S. Geng, G. de Melo, and Y. Zhang. Popcorn: Human-in-the-loop popularity debiasing in conversational recommender systems. In CIKM, pages 494--503, 2021.
[17]
N. Goyal, R. Paneri, A. Agarwal, U. Kalani, A. Sancheti, and N. Chhaya. Cam-gen: Causally aware metric-guided text generation. In ACL, pages 2047--2060, 2022.
[18]
A. Gruson, P. Chandar, C. Charbuillet, J. McInerney, S. Hansen, D. Tardieu, and B. Carterette. Offline evaluation to make decisions about playlist recommendation algorithms. In WSDM, pages 420--428. ACM, 2019.
[19]
P. Gupta, A. Sharma, P. Malhotra, L. Vig, and G. Shroff. Causer: Causal session-based recommendations for handling popularity bias. In CIKM, pages 3048--3052, 2021.
[20]
X. He, K. Deng, X. Wang, Y. Li, Y. Zhang, and M. Wang. Lightgcn: Simplifying and powering graph convolution network for recommendation. In SIGIR, pages 639--648, 2020.
[21]
X. He, L. Liao, H. Zhang, L. Nie, X. Hu, and T. Chua. Neural collaborative filtering. In WWW, pages 173--182, 2017.
[22]
M. Höfler. Causal inference based on counterfactuals. BMC medical research methodology, 5(1):1--12, 2005.
[23]
Y. Ji, A. Sun, J. Zhang, and C. Li. A re-visit of the popularity baseline in recommender systems. In SIGIR, pages 1749--1752, 2020.
[24]
T. Kamishima, S. Akaho, H. Asoh, and J. Sakuma. Correcting popularity bias by enhancing recommendation neutrality. In RecSys, volume 1247, 2014.
[25]
T. N. Kipf and M. Welling. Semi-supervised classification with graph convolutional networks. CoRR, abs/1609.02907, 2016.
[26]
S. Ko and J. Lee. User preference mining through collaborative filtering and content based filtering in recommender system. In EC-Web, volume 2455, pages 244--253, 2002.
[27]
D. Kowald, G. Mayr, M. Schedl, and E. Lex. A study on accuracy, miscalibration, and popularity bias in recommendations. CoRR, abs/2303.00400, 2023.
[28]
O. Lesota, A. B. Melchiorre, N. Rekabsaz, S. Brandl, D. Kowald, E. Lex, and M. Schedl. Analyzing item popularity bias of music recommender systems: Are different genders equally affected? In RecSys, pages 601--606, 2021.
[29]
X. Li and P. Li. Rejection sampling for weighted jaccard similarity revisited. In AAAI, pages 4197--4205, 2021.
[30]
A. Lin, J. Wang, Z. Zhu, and J. Caverlee. Quantifying and mitigating popularity bias in conversational recommender systems. In CIKM, pages 1238--1247, 2022.
[31]
F. Liu, Z. Cheng, L. Zhu, Z. Gao, and L. Nie. Interest-aware message-passing GCN for recommendation. In WWW, pages 1296--1305, 2021.
[32]
K. Liu, F. Xue, X. He, D. Guo, and R. Hong. Joint multi-grained popularity-aware graph convolution collaborative filtering for recommendation. TCSS, 10(1):72--83, 2023.
[33]
W. Ning, R. Cheng, J. Shen, N. A. H. Haldar, B. Kao, X. Yan, N. Huo, W. K. Lam, T. Li, and B. Tang. Automatic meta-path discovery for effective graph-based recommendation. In CIKM, pages 1563--1572, 2022.
[34]
W. Ning, X. Yan, W. Liu, R. Cheng, R. Zhang, and B. Tang. Multi-domain recommendation with embedding disentangling and domain alignment. In CIKM, 2023.
[35]
Y. Niu, K. Tang, H. Zhang, Z. Lu, X. Hua, and J. Wen. Counterfactual VQA: A cause-effect look at language bias. In CVPR, pages 12700--12710, 2021.
[36]
R. Otunba, R. A. Rufai, and J. Lin. MPR: multi-objective pairwise ranking. In RecSys, pages 170--178, 2017.
[37]
J. Pearl. Causality. Cambridge University Press, 2009.
[38]
C. Qian, F. Feng, L. Wen, C. Ma, and P. Xie. Counter-factual inference for text classification debiasing. In ACL/IJCNLP, pages 5434--5445, 2021.
[39]
S. Rendle, C. Freudenthaler, Z. Gantner, and L. Schmidt-Thieme. BPR: bayesian personalized ranking from implicit feedback. CoRR, abs/1205.2618, 2012.
[40]
T. Schnabel, A. Swaminathan, A. Singh, N. Chandak, and T. Joachims. Recommendations as treatments: Debiasing learning and evaluation. In ICML, volume 48 of JMLR Workshop and Conference Proceedings, pages 1670--1679, 2016.
[41]
W. Sun, S. Khenissi, O. Nasraoui, and P. Shafto. Debiasing the human-recommender system feedback loop in collaborative filtering. In WWW, pages 645--651, 2019.
[42]
K. Tang, Y. Niu, J. Huang, J. Shi, and H. Zhang. Unbiased scene graph generation from biased training. In CVPR, pages 3713--3722, 2020.
[43]
S. Tang, Q. Li, D. Wang, C. Gao, W. Xiao, D. Zhao, Y. Jiang, Q. Ma, and A. Zhang. Counterfactual video recommendation for duration debiasing. In SIGKDD, pages 4894--4903, 2023.
[44]
P. Velickovic, G. Cucurull, A. Casanova, A. Romero, P. Liò, and Y. Bengio. Graph attention networks. In ICLR, 2018.
[45]
Q. Wan, X. He, X. Wang, J. Wu, W. Guo, and R. Tang. Cross pairwise ranking for unbiased item recommendation. In WWW, pages 2370--2378, 2022.
[46]
J. Wang, Y. Chen, Z. Wang, and W. Zhao. Popularity-enhanced news recommendation with multi-view interest representation. In CIKM, pages 1949--1958, 2021.
[47]
W. Wang, F. Feng, X. He, H. Zhang, and T. S. Chua. Clicks can be cheating: Counterfactual recommendation for mitigating clickbait issue. In SIGIR, pages 1288--1297, 2021.
[48]
W. Wang, X. Lin, F. Feng, X. He, M. Lin, and T. S. Chua. Causal representation learning for out-of-distribution recommendation. In WWW, pages 3562--3571, 2022.
[49]
T. Wei, F. Feng, J. Chen, Z. Wu, J. Yi, and X. He. Model-agnostic counterfactual reasoning for eliminating popularity bias in recommender system. In SIGKDD, pages 1791--1800, 2021.
[50]
Y. Wu and X. Huang. A gumbel-based rating prediction framework for imbalanced recommendation. In CIKM, pages 2199--2209, 2022.
[51]
G. Xv, C. Lin, H. Li, J. Su, W. Ye, and Y. Chen. Neutralizing popularity bias in recommendation models. In SIGIR, pages 2623--2628, 2022.
[52]
X. Yang, F. Feng, W. Ji, M. Wang, and T. Chua. Deconfounded video moment retrieval with causal intervention. In SIGIR, pages 1--10, 2021.
[53]
Q. Zhang, X. Zhang, Y. Liu, H. Wang, M. Gao, J. Zhang, and R. Guo. Debiasing recommendation by learning identifiable latent confounders. In SIGKDD, pages 3353--3363, 2023.
[54]
Y. Zhang, F. Feng, X. He, T. Wei, C. Song, G. Ling, and Y. Zhang. Causal intervention for leveraging popularity bias in recommendation. In SIGIR, pages 11--20, 2021.
[55]
Y. Zhang, W. Wang, P. Wu, F. Feng, and X. He. Causal recommendation: Progresses and future directions. In WWW, 2022.
[56]
M. Zhao, L. Wu, Y. Liang, L. Chen, J. Zhang, Q. Deng, K. Wang, X. Shen, T. Lv, and R. Wu. Investigating accuracy-novelty performance for graph-based collaborative filtering. In SIGIR, pages 50--59, 2022.
[57]
Y. Zheng, C. Gao, X. Li, X. He, Y. Li, and D. Jin. Disentangling user interest and conformity for recommendation with causal embedding. In WWW, pages 2980--2991, 2021.
[58]
X. Zhu, Y. Zhang, F. Feng, X. Yang, D. Wang, and X. He. Mitigating hidden confounding effects for causal recommendation. CoRR, abs/2205.07499, 2022.
[59]
Y. Zhu, J. Yi, J. Xie, and Z. Chen. Deep causal reasoning for recommendations. CoRR, abs/2201.02088, 2022.
[60]
Z. Zhu, Y. He, X. Zhao, Y. Zhang, J. Wang, and J. Caverlee. Popularity-opportunity bias in collaborative filtering. In WSDM, pages 85--93, 2021.
Index Terms
- Debiasing Recommendation with Personal Popularity
Recommendations
The Unfairness of Popularity Bias in Music Recommendation: A Reproducibility Study
Advances in Information RetrievalAbstractResearch has shown that recommender systems are typically biased towards popular items, which leads to less popular items being underrepresented in recommendations. The recent work of Abdollahpouri et al. in the context of movie recommendations ...
An Enhanced Collaborative Filtering with Flexible Item Popularity Control for Recommender Systems
SocialCom '14: Proceedings of the 2014 International Conference on Social ComputingWith the emerging and rapid development of Internet applications like social networks, E-commerce and so on, massive information has been created and stored. Recommender systems have been developed to deal with the information overload problem. Various ...
An Enhanced Collaborative Filtering with Flexible Item Popularity Control for Recommender Systems
SocialCom '14: Proceedings of the 2014 International Conference on Social ComputingWith the emerging and rapid development of Internet applications like social networks, E-commerce and so on, massive information has been created and stored. Recommender systems have been developed to deal with the information overload problem. Various ...
Comments
Information & Contributors
Information
Published In
May 2024
4826 pages
ISBN:9798400701719
DOI:10.1145/3589334
- General Chairs:
- Tat-Seng Chua,
- Chong-Wah Ngo,
- Proceedings Chair:
- Roy Ka-Wei Lee,
- Program Chairs:
- Ravi Kumar,
- Hady W. Lauw
Copyright © 2024 ACM.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].
Sponsors
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 13 May 2024
Check for updates
Badges
Author Tags
Qualifiers
- Research-article
Data Availability
Video presentation https://dl.acm.org/doi/10.1145/3589334.3645421#Debiasing Recommendation with Personal Popularity.mp4
Supplemental video https://dl.acm.org/doi/10.1145/3589334.3645421#rfp0610.mp4
Funding Sources
- the University of Hong Kong
- the HKU Outstanding Research Student Supervisor Award 2022-23
- Shenzhen Fundamental Research Program
- the France/Hong Kong Joint Research Scheme 2020/21
- the Guangdong Provincial Key Laboratory
- the Hong Kong Jockey Club Charities Trust
Conference
WWW '24
Sponsor:
Acceptance Rates
Overall Acceptance Rate 1,899 of 8,196 submissions, 23%
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 106Total Downloads
- Downloads (Last 12 months)106
- Downloads (Last 6 weeks)52
Other Metrics
Citations
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in