Sequential Recommendation
Sequential recommendation aims to predict users' next interactions by analyzing their historical behavior sequences, focusing on capturing dynamic preferences and long-term patterns. Current research heavily utilizes transformer-based architectures, large language models (LLMs), and graph neural networks (GNNs), often incorporating techniques like contrastive learning, test-time training, and knowledge distillation to improve accuracy and efficiency, particularly for large-scale datasets. This field is crucial for personalized recommendations in various applications, driving improvements in user experience and business outcomes through more accurate and timely predictions. Addressing challenges like scalability, noise in user data, and the effective integration of diverse data sources (e.g., textual item descriptions, collaborative filtering signals) remains a key focus.
Papers
Model-enhanced Contrastive Reinforcement Learning for Sequential Recommendation
Chengpeng Li, Zhengyi Yang, Jizhi Zhang, Jiancan Wu, Dingxian Wang, Xiangnan He, Xiang Wang
Multiple Key-value Strategy in Recommendation Systems Incorporating Large Language Model
Dui Wang, Xiangyu Hou, Xiaohui Yang, Bo Zhang, Renbing Chen, Daiyue Xue
LlamaRec: Two-Stage Recommendation using Large Language Models for Ranking
Zhenrui Yue, Sara Rabhi, Gabriel de Souza Pereira Moreira, Dong Wang, Even Oldridge