Retrieval Augmented Conversational Recommendation with Reinforcement Learning

TL;DR

This paper introduces RAR, a retrieval-augmented conversational recommendation framework that combines retrieval, generation, and reinforcement learning to improve recommendation accuracy and factuality in movie recommendations.

Contribution

The paper presents RAR, a novel two-stage retrieval and generation framework with reinforcement learning, and constructs a large-scale movie corpus for enhanced CRS performance.

Findings

RAR outperforms state-of-the-art methods on multiple benchmarks.

The reinforcement learning approach improves retrieval relevance and reduces hallucinations.

Grounding in factual metadata enhances recommendation accuracy.

Abstract

Large language models (LLMs) exhibit enhanced capabilities in language understanding and generation. By utilizing their embedded knowledge, LLMs are increasingly used as conversational recommender systems (CRS), achieving improved performance across diverse scenarios. However, existing LLM-based methods rely on pretrained knowledge without external retrieval mechanisms for novel items. Additionally, the lack of a unified corpus poses challenges for integrating retrieval augmentation into CRS. Motivated by these challenges, we present RAR, a novel two-stage retrieval augmented conversational recommendation framework that aligns retrieval and generation to enhance both performance and factuality. To support this framework and provide a unified corpus, we construct a large-scale movie corpus, comprising over 300k movies with rich metadata, such as titles, casts and plot summaries. Leveraging this data, our primary contribution is RAR, the first framework to departs from standard two-stage CRS by dynamically bridging retrieval and generation. First, a retriever model generates candidate items based on user history; in the subsequent stage, an LLM refines the recommendations by incorporating conversational context with retrieved results. In addition, we introduce a novel reinforcement learning (RL) method that leverages LLM feedback to iteratively update the retriever. By creating a collaborative feedback loop that reinforces sampled candidate sets with higher ranking metrics, RAR effectively mitigates the misalignment between the retrieval and generation stages. Furthermore, grounding the LLM in factual metadata allows our RL-driven approach to capture subtle user intentions and generate context-aware recommendations with reduced hallucinations. We validate our approach through extensive experiments on multiple benchmarks, where RAR consistently outperforms state-of-the-art baseline methods.