Interplay: Training Independent Simulators for Reference-Free Conversational Recommendation

TL;DR

This paper introduces a novel reference-free simulation framework for training conversational recommender systems using two independent LLMs, resulting in more realistic and diverse dialogues without relying on pre-defined target items.

Contribution

The authors propose a new reference-free simulation approach with two independent LLMs that interact dynamically, improving realism and scalability in training conversational recommenders.

Findings

Simulators produce more realistic, diverse conversations.

Approach matches or exceeds existing methods in quality.

Scalable data generation without pre-defined target items.

Abstract

Training conversational recommender systems (CRS) requires extensive dialogue data, which is challenging to collect at scale. To address this, researchers have used simulated user-recommender conversations. Traditional simulation approaches often utilize a single large language model (LLM) that generates entire conversations with prior knowledge of the target items, leading to scripted and artificial dialogues. We propose a reference-free simulation framework that trains two independent LLMs, one as the user and one as the conversational recommender. These models interact in real-time without access to predetermined target items, but preference summaries and target attributes, enabling the recommender to genuinely infer user preferences through dialogue. This approach produces more realistic and diverse conversations that closely mirror authentic human-AI interactions. Our reference-free simulators match or exceed existing methods in quality, while offering a scalable solution for generating high-quality conversational recommendation data without constraining conversations to pre-defined target items. We conduct both quantitative and human evaluations to confirm the effectiveness of our reference-free approach.