Explorando Formas de Calibração e Redução do Viés de Popularidade em Sistemas de Recomendação
Resumo
Recommender systems aim to enhance user engagement by suggesting new content based on their interests. Common algorithms like collaborative filtering recommend items similar to those users prefer, facilitating content discovery. However, recent research has identified issues such as unfairness, calibration errors, and popularity bias. While popularity bias promotes popular item consumption, it can lead to unfair recommendations that do not accurately reflect individual user interests. Current calibration methods focus on fairness but often overlook the amplifying effect of popularity bias. Our study addresses this gap by investigating calibration techniques and bias reduction strategies to deliver fairer recommendations aligned with user preferences, across various popularity levels. This research contributes insights into system calibration, fairness, user experience, and evaluation metrics.
Palavras-chave:
Sistemas de Recomendação, Viés de Popularidade, Justiça, Calibração
Referências
Himan Abdollahpouri, Robin Burke, and Bamshad Mobasher. 2019. Managing popularity bias in recommender systems with personalized re-ranking. In The thirty-second international flairs conference. AAAI Press, California, USA.
Jiawei Chen, Hande Dong, Xiang Wang, Fuli Feng, Meng Wang, and Xiangnan He. 2023. Bias and debias in recommender system: A survey and future directions. ACM Transactions on Information Systems 41, 3 (2023), 1–39.
Zhihong Chen, Jiawei Wu, Chenliang Li, Jingxu Chen, Rong Xiao, and Binqiang Zhao. 2022. Co-training disentangled domain adaptation network for leveraging popularity bias in recommenders. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. ACM, New York, NY, USA, 60–69.
Diego Corrêa da Silva, Marcelo Garcia Manzato, and Frederico Araújo Durão. 2021. Exploiting personalized calibration and metrics for fairness recommendation. Expert Systems with Applications 181 (2021), 115112.
Oleg Lesota, Alessandro Melchiorre, Navid Rekabsaz, Stefan Brandl, Dominik Kowald, Elisabeth Lex, and Markus Schedl. 2021. Analyzing item popularity bias of music recommender systems: are different genders equally affected?. In Proceedings of the 15th ACM Conference on Recommender Systems. ACM, New York, NY, USA, 601–606.
Allen Lin, Jianling Wang, Ziwei Zhu, and James Caverlee. 2022. Quantifying and mitigating popularity bias in conversational recommender systems. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management. ACM, New York, NY, USA, 1238–1247.
Mohammadmehdi Naghiaei, Hossein A Rahmani, and Mahdi Dehghan. 2022. The unfairness of popularity bias in book recommendation. In International Workshop on Algorithmic Bias in Search and Recommendation. Springer, New York, NY, USA, 69–81.
Steffen Rendle, Christoph Freudenthaler, Zeno Gantner, and Lars Schmidt-Thieme. 2012. BPR: Bayesian personalized ranking from implicit feedback. arXiv preprint arXiv:1205.2618 (2012).
Andrê Sacilotti, Marcelo Garcia Manzato, and Rodrigo Ferrari de Souza. 2022. Counteracting popularity-bias and improving diversity through calibrated recommendations. In Sixteenth ACM conference on recommender systems. Association for Computing Machinery, New York, NY, USA. (Sob avaliação).
Rodrigo Souza and Marcelo Manzato. 2024. A Two-Stage Calibration Approach for Mitigating Bias and Fairness in Recommender Systems. In Proceedings of the 39th ACM/SIGAPP Symposium on Applied Computing. ACM, New York, NY, USA, 1659–1661.
Harald Steck. 2018. Calibrated Recommendations. In Proceedings of the 12th ACM Conference on Recommender Systems (Vancouver, British Columbia, Canada) (RecSys ’18). Association for Computing Machinery, New York, NY, USA, 154–162. DOI: 10.1145/3240323.3240372
Emre Yalcin. 2021. Blockbuster: A new perspective on popularity-bias in recommender systems. In 2021 6th International Conference on Computer Science and Engineering (UBMK). IEEE, New York, NY, USA, 107–112.
Emre Yalcin and Alper Bilge. 2022. Treating adverse effects of blockbuster bias on beyond-accuracy quality of personalized recommendations. Engineering Science and Technology, an International Journal 33 (2022), 101083
Jiawei Chen, Hande Dong, Xiang Wang, Fuli Feng, Meng Wang, and Xiangnan He. 2023. Bias and debias in recommender system: A survey and future directions. ACM Transactions on Information Systems 41, 3 (2023), 1–39.
Zhihong Chen, Jiawei Wu, Chenliang Li, Jingxu Chen, Rong Xiao, and Binqiang Zhao. 2022. Co-training disentangled domain adaptation network for leveraging popularity bias in recommenders. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. ACM, New York, NY, USA, 60–69.
Diego Corrêa da Silva, Marcelo Garcia Manzato, and Frederico Araújo Durão. 2021. Exploiting personalized calibration and metrics for fairness recommendation. Expert Systems with Applications 181 (2021), 115112.
Oleg Lesota, Alessandro Melchiorre, Navid Rekabsaz, Stefan Brandl, Dominik Kowald, Elisabeth Lex, and Markus Schedl. 2021. Analyzing item popularity bias of music recommender systems: are different genders equally affected?. In Proceedings of the 15th ACM Conference on Recommender Systems. ACM, New York, NY, USA, 601–606.
Allen Lin, Jianling Wang, Ziwei Zhu, and James Caverlee. 2022. Quantifying and mitigating popularity bias in conversational recommender systems. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management. ACM, New York, NY, USA, 1238–1247.
Mohammadmehdi Naghiaei, Hossein A Rahmani, and Mahdi Dehghan. 2022. The unfairness of popularity bias in book recommendation. In International Workshop on Algorithmic Bias in Search and Recommendation. Springer, New York, NY, USA, 69–81.
Steffen Rendle, Christoph Freudenthaler, Zeno Gantner, and Lars Schmidt-Thieme. 2012. BPR: Bayesian personalized ranking from implicit feedback. arXiv preprint arXiv:1205.2618 (2012).
Andrê Sacilotti, Marcelo Garcia Manzato, and Rodrigo Ferrari de Souza. 2022. Counteracting popularity-bias and improving diversity through calibrated recommendations. In Sixteenth ACM conference on recommender systems. Association for Computing Machinery, New York, NY, USA. (Sob avaliação).
Rodrigo Souza and Marcelo Manzato. 2024. A Two-Stage Calibration Approach for Mitigating Bias and Fairness in Recommender Systems. In Proceedings of the 39th ACM/SIGAPP Symposium on Applied Computing. ACM, New York, NY, USA, 1659–1661.
Harald Steck. 2018. Calibrated Recommendations. In Proceedings of the 12th ACM Conference on Recommender Systems (Vancouver, British Columbia, Canada) (RecSys ’18). Association for Computing Machinery, New York, NY, USA, 154–162. DOI: 10.1145/3240323.3240372
Emre Yalcin. 2021. Blockbuster: A new perspective on popularity-bias in recommender systems. In 2021 6th International Conference on Computer Science and Engineering (UBMK). IEEE, New York, NY, USA, 107–112.
Emre Yalcin and Alper Bilge. 2022. Treating adverse effects of blockbuster bias on beyond-accuracy quality of personalized recommendations. Engineering Science and Technology, an International Journal 33 (2022), 101083
Publicado
14/10/2024
Como Citar
SOUZA, Rodrigo F.; MANZATO, Marcelo G..
Explorando Formas de Calibração e Redução do Viés de Popularidade em Sistemas de Recomendação. In: CONCURSO DE TESES E DISSERTAÇÕES - SIMPÓSIO BRASILEIRO DE SISTEMAS MULTIMÍDIA E WEB (WEBMEDIA), 30. , 2024, Juiz de Fora/MG.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2024
.
p. 9-10.
ISSN 2596-1683.
DOI: https://doi.org/10.5753/webmedia_estendido.2024.244380.
