KELLMRec: Knowledge-Enhanced Large Language Models for Recommendation

• DOI: 10.48550/arxiv.2403.06642
• Publication Date: 2024-03-11
• Authors: Weiqing Luo; Chonggang Song; Lingling Yi; Gong Cheng
• Keywords: FOS: Computer and information sciences; Computer Science - Computation and Language; Artificial Intelligence (cs.AI); Computer Science - Artificial Intelligence; Computation and Language (cs.CL); Information Retrieval (cs.IR); Computer Science - Information Retrieval

The utilization of semantic information is an important research problem in the field recommender systems, which aims to complement missing parts mainstream ID-based approaches. With rise LLM, its ability act as a knowledge base and reasoning capability have opened up new possibilities for this area, making LLM-based recommendation emerging direction. However, directly using LLM process scenarios unreliable sub-optimal due several problems such hallucination. A promising way cope with use external aid generating truthful usable text. Inspired by above motivation, we propose Knowledge-Enhanced LLMRec method. In addition prompts, proposed method also includes knowledge-based contrastive learning scheme training. Experiments on public datasets in-enterprise validate effectiveness