Abstract
Leveraging Large Language Models (LLMs) to harness user-item interaction histories for item generation has emerged as a promising paradigm in generative recommendation. However, the limited context window of LLMs often restricts them to focusing on recent user interactions only, leading to the neglect of long-term interests involved in the longer histories. To address this challenge, we propose a novel extit{Automatic Memory-Retrieval} framework, extbf{ extit{AutoMR}}, which is capable of storing long-term interests in the memory and extracting relevant information from it for next-item generation within LLMs. Extensive experimental results on two real-world datasets demonstrate the effectiveness of our proposed AutoMR framework in utilizing long-term interests for generative recommendation.
Citation:
@article{wang2024leveraging,
title={Leveraging Memory Retrieval to Enhance LLM-based Generative Recommendation},
author={Wang, Chengbing and Zhang, Yang and Zhu, Fengbin and Zhang, Jizhi and Shi, Tianhao and Feng, Fuli},
journal={arXiv preprint arXiv:2412.17593},
year={2024}
}
Chengbing Wang
王城冰
Jizhi Zhang
张及之
Tianhao Shi
史天昊
