作者位次 #5/7SIGIR 2026Oral· Information Retrieval (Full)
Divergence Meets Consensus: A Multi-Source Negative Sampling Framework for Sequential Recommendation
Yuanzi Li · Lingjie Wang · Jingyu Zhao · Zihang Tian · Yuhan Wang · Lei Wang · Xu Chen
摘要
Negative sampling is significant for training sequential recommendation models under implicit feedback. The predominant strategy, self-guided hard negative sampling, selects negatives based on the model’s current state but suffers from three limitations: (1) the coupling between sampling and mode...