Enhancing High-order Interaction Awareness in LLM-based Recommender Model

TL;DR

This paper introduces ELMRec, an improved LLM-based recommender that enhances high-order interaction modeling and reranking, leading to superior recommendation performance without pre-training on graphs.

Contribution

The paper proposes a novel method to better model user-item interactions in LLM recommenders using enhanced whole-word embeddings and a reranking strategy.

Findings

ELMRec outperforms state-of-the-art methods in direct recommendations.

Enhanced embeddings improve interpretation of graph-constructed interactions.

Reranking based on recent interactions improves recommendation relevance.

Abstract

Large language models (LLMs) have demonstrated prominent reasoning capabilities in recommendation tasks by transforming them into text-generation tasks. However, existing approaches either disregard or ineffectively model the user-item high-order interactions. To this end, this paper presents an enhanced LLM-based recommender (ELMRec). We enhance whole-word embeddings to substantially enhance LLMs' interpretation of graph-constructed interactions for recommendations, without requiring graph pre-training. This finding may inspire endeavors to incorporate rich knowledge graphs into LLM-based recommenders via whole-word embedding. We also found that LLMs often recommend items based on users' earlier interactions rather than recent ones, and present a reranking solution. Our ELMRec outperforms state-of-the-art (SOTA) methods in both direct and sequential recommendations.