Relieving popularity bias in recommendation via debiasing representation enhancement
Abstract The interaction data used for training recommender systems often exhibit a long-tail distribution. Such highly imbalanced data distribution results in an unfair learning process among items. Contrastive learning alleviates the above issue by data augmentation. However, it lacks consideratio...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2024-11-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-024-01649-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571194656686080 |
|---|---|
| author | Junsan Zhang Sini Wu Te Wang Fengmei Ding Jie Zhu |
| author_facet | Junsan Zhang Sini Wu Te Wang Fengmei Ding Jie Zhu |
| author_sort | Junsan Zhang |
| collection | DOAJ |
| description | Abstract The interaction data used for training recommender systems often exhibit a long-tail distribution. Such highly imbalanced data distribution results in an unfair learning process among items. Contrastive learning alleviates the above issue by data augmentation. However, it lacks consideration of the significant disparity in popularity between items and may even introduce false negatives during the data augmentation, misleading user preference prediction. To address this issue, we combine contrastive learning with a weighted model for negative validation. By penalizing identified false negatives during training, we limit their potential harm within the training process. Meanwhile, to tackle the scarcity of supervision signals for unpopular items, we design Popularity Associated Modeling to mine the correlation among items. Then we guide unpopular items to learn hidden features favored by specific users from their associated popular items, which provides effective supplementary information for their representation modeling. Extensive experiments on three real-world datasets demonstrate that our proposed model outperforms state-of-the-art baselines in recommendation performance, with Recall@20 improvements of 4.2%, 2.4% and 3.6% across the datasets, but also shows significant effectiveness in relieving popularity bias. |
| format | Article |
| id | doaj-art-c01a3b549d8344409244fa6fcf9620be |
| institution | Kabale University |
| issn | 2199-4536 2198-6053 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj-art-c01a3b549d8344409244fa6fcf9620be2025-02-02T12:50:14ZengSpringerComplex & Intelligent Systems2199-45362198-60532024-11-0111111410.1007/s40747-024-01649-zRelieving popularity bias in recommendation via debiasing representation enhancementJunsan Zhang0Sini Wu1Te Wang2Fengmei Ding3Jie Zhu4College of Computer Science and Technology, China University of Petroleum (East China)College of Computer Science and Technology, China University of Petroleum (East China)College of Computer Science and Technology, China University of Petroleum (East China)College of Computer Science and Technology, China University of Petroleum (East China)College of Mathematics and Information Science, Hebei UniversityAbstract The interaction data used for training recommender systems often exhibit a long-tail distribution. Such highly imbalanced data distribution results in an unfair learning process among items. Contrastive learning alleviates the above issue by data augmentation. However, it lacks consideration of the significant disparity in popularity between items and may even introduce false negatives during the data augmentation, misleading user preference prediction. To address this issue, we combine contrastive learning with a weighted model for negative validation. By penalizing identified false negatives during training, we limit their potential harm within the training process. Meanwhile, to tackle the scarcity of supervision signals for unpopular items, we design Popularity Associated Modeling to mine the correlation among items. Then we guide unpopular items to learn hidden features favored by specific users from their associated popular items, which provides effective supplementary information for their representation modeling. Extensive experiments on three real-world datasets demonstrate that our proposed model outperforms state-of-the-art baselines in recommendation performance, with Recall@20 improvements of 4.2%, 2.4% and 3.6% across the datasets, but also shows significant effectiveness in relieving popularity bias.https://doi.org/10.1007/s40747-024-01649-zRecommender systemPopularity biasCollaborative filteringContrastive learning |
| spellingShingle | Junsan Zhang Sini Wu Te Wang Fengmei Ding Jie Zhu Relieving popularity bias in recommendation via debiasing representation enhancement Complex & Intelligent Systems Recommender system Popularity bias Collaborative filtering Contrastive learning |
| title | Relieving popularity bias in recommendation via debiasing representation enhancement |
| title_full | Relieving popularity bias in recommendation via debiasing representation enhancement |
| title_fullStr | Relieving popularity bias in recommendation via debiasing representation enhancement |
| title_full_unstemmed | Relieving popularity bias in recommendation via debiasing representation enhancement |
| title_short | Relieving popularity bias in recommendation via debiasing representation enhancement |
| title_sort | relieving popularity bias in recommendation via debiasing representation enhancement |
| topic | Recommender system Popularity bias Collaborative filtering Contrastive learning |
| url | https://doi.org/10.1007/s40747-024-01649-z |
| work_keys_str_mv | AT junsanzhang relievingpopularitybiasinrecommendationviadebiasingrepresentationenhancement AT siniwu relievingpopularitybiasinrecommendationviadebiasingrepresentationenhancement AT tewang relievingpopularitybiasinrecommendationviadebiasingrepresentationenhancement AT fengmeiding relievingpopularitybiasinrecommendationviadebiasingrepresentationenhancement AT jiezhu relievingpopularitybiasinrecommendationviadebiasingrepresentationenhancement |