Triplet Contrastive Learning with Learnable Sequence Augmentation for Sequential Recommendation
The quality of augmented data directly affects the performance of contrastive learning. Low-quality augmentation offers limited benefits for model optimization. Existing contrastive learning-based sequential recommendation works primarily utilize heuristic data augmentation methods, which often exhibit excessive randomness and struggle to generate positive samples that align with users' true intentions.
Authors
- Xianye Ben (ORCID: https://orcid.org/0000-0001-8083-3501)
- Pengjie Ren (ORCID: https://orcid.org/0000-0003-2964-6422)
- Yujie Lin (ORCID: https://orcid.org/0000-0002-2146-0626)
- Jianli Zhao (ORCID: https://orcid.org/0000-0002-7291-9003)
- Moyan Zhang (ORCID: https://orcid.org/0000-0001-6130-1286)
- Wei Wang (ORCID: https://orcid.org/0000-0002-7080-3381)
- Jie Sun
- Yujun Li (ORCID: https://orcid.org/0000-0003-4455-5991)
- Hongyu Lu (ORCID: https://orcid.org/0000-0002-0247-2496)
Institutions
- Shandong University (CN)
- Tencent (China) (CN)
- Shandong Provincial Hospital (CN)
- Zhejiang Lab (CN)
- Shandong University of Science and Technology (CN)
Publication Details
- Published
- 2025-07-13
- DOI
- https://doi.org/10.1145/3726302.3730101
- Primary Topic
- Recommender Systems and Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Texas Space Grant Consortium
- Tencent
- National Natural Science Foundation of China