Towards Efficient Hypergraph and Multi-LLM Agent Recommender Systems

TL;DR

This paper introduces HGLMRec, a hypergraph-based multi-LLM agent recommender system that improves recommendation accuracy and reduces computational costs by capturing complex user-item relationships and optimizing token retrieval.

Contribution

The paper presents a novel hypergraph encoder integrated with multi-LLM agents for recommender systems, addressing hallucination and efficiency issues in generative recommendation methods.

Findings

Performance surpasses state-of-the-art baselines.

Reduces computational overhead during inference.

Enhances recommendation quality with complex relationship modeling.

Abstract

Recommender Systems (RSs) have become the cornerstone of various applications such as e-commerce and social media platforms. The evolution of RSs is paramount in the digital era, in which personalised user experience is tailored to the user's preferences. Large Language Models (LLMs) have sparked a new paradigm - generative retrieval and recommendation. Despite their potential, generative RS methods face issues such as hallucination, which degrades the recommendation performance, and high computational cost in practical scenarios. To address these issues, we introduce HGLMRec, a novel Multi-LLM agent-based RS that incorporates a hypergraph encoder designed to capture complex, multi-behaviour relationships between users and items. The HGLMRec model retrieves only the relevant tokens during inference, reducing computational overhead while enriching the retrieval context. Experimental results show performance improvement by HGLMRec against state-of-the-art baselines at lower computational cost.