Taming Recommendation Bias with Causal Intervention on Evolving Personal Popularity
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910972390670336 |
|---|---|
| author | Tan, Shiyin Li, Dongyuan Jiang, Renhe Wang, Zhen Yu, Xingtong Okumura, Manabu |
| author_facet | Tan, Shiyin Li, Dongyuan Jiang, Renhe Wang, Zhen Yu, Xingtong Okumura, Manabu |
| contents | Popularity bias occurs when popular items are recommended far more frequently than they should be, negatively impacting both user experience and recommendation accuracy. Existing debiasing methods mitigate popularity bias often uniformly across all users and only partially consider the time evolution of users or items. However, users have different levels of preference for item popularity, and this preference is evolving over time. To address these issues, we propose a novel method called CausalEPP (Causal Intervention on Evolving Personal Popularity) for taming recommendation bias, which accounts for the evolving personal popularity of users. Specifically, we first introduce a metric called {Evolving Personal Popularity} to quantify each user's preference for popular items. Then, we design a causal graph that integrates evolving personal popularity into the conformity effect, and apply deconfounded training to mitigate the popularity bias of the causal graph. During inference, we consider the evolution consistency between users and items to achieve a better recommendation. Empirical studies demonstrate that CausalEPP outperforms baseline methods in reducing popularity bias while improving recommendation accuracy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_14310 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Taming Recommendation Bias with Causal Intervention on Evolving Personal Popularity Tan, Shiyin Li, Dongyuan Jiang, Renhe Wang, Zhen Yu, Xingtong Okumura, Manabu Information Retrieval Machine Learning Popularity bias occurs when popular items are recommended far more frequently than they should be, negatively impacting both user experience and recommendation accuracy. Existing debiasing methods mitigate popularity bias often uniformly across all users and only partially consider the time evolution of users or items. However, users have different levels of preference for item popularity, and this preference is evolving over time. To address these issues, we propose a novel method called CausalEPP (Causal Intervention on Evolving Personal Popularity) for taming recommendation bias, which accounts for the evolving personal popularity of users. Specifically, we first introduce a metric called {Evolving Personal Popularity} to quantify each user's preference for popular items. Then, we design a causal graph that integrates evolving personal popularity into the conformity effect, and apply deconfounded training to mitigate the popularity bias of the causal graph. During inference, we consider the evolution consistency between users and items to achieve a better recommendation. Empirical studies demonstrate that CausalEPP outperforms baseline methods in reducing popularity bias while improving recommendation accuracy. |
| title | Taming Recommendation Bias with Causal Intervention on Evolving Personal Popularity |
| topic | Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2505.14310 |