Saved in:
Bibliographic Details
Main Authors: Wang, Xinfeng, Cui, Jin, Suzuki, Yoshimi, Fukumoto, Fumiyo
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.10587
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910776434884608
author Wang, Xinfeng
Cui, Jin
Suzuki, Yoshimi
Fukumoto, Fumiyo
author_facet Wang, Xinfeng
Cui, Jin
Suzuki, Yoshimi
Fukumoto, Fumiyo
contents Large language model (LLM)-based recommender models that bridge users and items through textual prompts for effective semantic reasoning have gained considerable attention. However, few methods consider the underlying rationales behind interactions, such as user preferences and item attributes, limiting the reasoning capability of LLMs for recommendations. This paper proposes a rationale distillation recommender (RDRec), a compact model designed to learn rationales generated by a larger language model (LM). By leveraging rationales from reviews related to users and items, RDRec remarkably specifies their profiles for recommendations. Experiments show that RDRec achieves state-of-the-art (SOTA) performance in both top-N and sequential recommendations. Our source code is released at https://github.com/WangXFng/RDRec.
format Preprint
id arxiv_https___arxiv_org_abs_2405_10587
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle RDRec: Rationale Distillation for LLM-based Recommendation
Wang, Xinfeng
Cui, Jin
Suzuki, Yoshimi
Fukumoto, Fumiyo
Computation and Language
Information Retrieval
Large language model (LLM)-based recommender models that bridge users and items through textual prompts for effective semantic reasoning have gained considerable attention. However, few methods consider the underlying rationales behind interactions, such as user preferences and item attributes, limiting the reasoning capability of LLMs for recommendations. This paper proposes a rationale distillation recommender (RDRec), a compact model designed to learn rationales generated by a larger language model (LM). By leveraging rationales from reviews related to users and items, RDRec remarkably specifies their profiles for recommendations. Experiments show that RDRec achieves state-of-the-art (SOTA) performance in both top-N and sequential recommendations. Our source code is released at https://github.com/WangXFng/RDRec.
title RDRec: Rationale Distillation for LLM-based Recommendation
topic Computation and Language
Information Retrieval
url https://arxiv.org/abs/2405.10587