LLM2Rec: Large Language Models Are Powerful Embedding Models for Sequential Recommendation

TL;DR

LLM2Rec introduces a novel embedding model that combines the semantic understanding of large language models with collaborative filtering signals to enhance sequential recommendation performance across diverse domains.

Contribution

The paper presents a two-stage training framework for integrating LLMs with CF signals, improving recommendation accuracy and generalization beyond existing methods.

Findings

Significant improvement in recommendation quality on real-world datasets.

Enhanced out-of-domain recommendation performance.

Effective integration of semantic and collaborative information.

Abstract

Sequential recommendation aims to predict users' future interactions by modeling collaborative filtering (CF) signals from historical behaviors of similar users or items. Traditional sequential recommenders predominantly rely on ID-based embeddings, which capture CF signals through high-order co-occurrence patterns. However, these embeddings depend solely on past interactions, lacking transferable knowledge to generalize to unseen domains. Recent advances in large language models (LLMs) have motivated text-based recommendation approaches that derive item representations from textual descriptions. While these methods enhance generalization, they fail to encode CF signals-i.e., latent item correlations and preference patterns-crucial for effective recommendation. We argue that an ideal embedding model should seamlessly integrate CF signals with rich semantic representations to improve both in-domain and out-of-domain recommendation performance. To this end, we propose LLM2Rec, a novel embedding model tailored for sequential recommendation, integrating the rich semantic understanding of LLMs with CF awareness. Our approach follows a two-stage training framework: (1) Collaborative Supervised Fine-tuning, which adapts LLMs to infer item relationships based on historical interactions, and (2) Item-level Embedding Modeling, which refines these specialized LLMs into structured item embedding models that encode both semantic and collaborative information. Extensive experiments on real-world datasets demonstrate that LLM2Rec effectively improves recommendation quality across both in-domain and out-of-domain settings. Our findings highlight the potential of leveraging LLMs to build more robust, generalizable embedding models for sequential recommendation. Our codes are available at https://github.com/HappyPointer/LLM2Rec.