Synthetic Dialogue Generation for Interactive Conversational Elicitation & Recommendation (ICER)

2025/09/26 by Moonkyung Ryu, Ryu, Moonkyung, Chih‐Wei Hsu +7 · 1 citation
Computer Science · #AI in Service Interactions #Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #FOS: Computer and information sciences #Speech and dialogue systems #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2510.02331

openalex publication_date 2025/09/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

While language models (LMs) offer great potential for conversational recommender systems (CRSs), the paucity of public CRS data makes fine-tuning LMs for CRSs challenging. In response, LMs as user simulators qua data generators can be used to train LM-based CRSs, but often lack behavioral consistency, generating utterance sequences inconsistent with those of any real user. To address this, we develop a methodology for generating natural dialogues that are consistent with a user's underlying state using behavior simulators together with LM-prompting. We illustrate our approach by generating a large, open-source CRS data set with both preference elicitation and example critiquing. Rater evaluation on some of these dialogues shows them to exhibit considerable consistency, factuality and naturalness.

Citations

• The Challenge of Using LLMs to Simulate Human Behavior: A Causal Inference Perspective
• Preference Elicitation with Soft Attributes in Interactive Recommendation
• Demystifying Embedding Spaces using Large Language Models
• Generative Agents: Interactive Simulacra of Human Behavior
• GPT-4 Technical Report
• Talk the Walk: Synthetic Data Generation for Conversational Music Recommendation
• Evaluation of Synthetic Datasets for Conversational Recommender Systems
• Dialog Inpainting: Turning Documents into Dialogs
• Training language models to follow instructions with human feedback
• BNAI, NO-TOKEN, and MIND-UNITY: Pillars of a Systemic Revolution in Artificial Intelligence
• Towards Topic-Guided Conversational Recommender System
• Gradient-based Optimization for Bayesian Preference Elicitation
• Towards Scalable Multi-domain Conversational Agents: The Schema-Guided\n Dialogue Dataset
• Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)
• Deep reinforcement learning from human preferences
• Axiomatic Attribution for Deep Networks
• Evaluating Prerequisite Qualities for Learning End-to-End Dialog Systems
• Understanding the Metropolis-Hastings Algorithm
• Prefix-Tuning: Optimizing Continuous Prompts for Generation
• Prefix-Tuning: Optimizing Continuous Prompts for Generation

Cited by

• Context Engineering 2.0: The Context of Context Engineering

Related

• Leveraging Large Language Models in Conversational Recommender Systems
• Evaluating Large Language Models as Generative User Simulators for Conversational Recommendation
• A LLM-based Controllable, Scalable, Human-Involved User Simulator Framework for Conversational Recommender Systems
• How Reliable is Your Simulator? Analysis on the Limitations of Current LLM-based User Simulators for Conversational Recommendation
• RecUserSim: A Realistic and Diverse User Simulator for Evaluating Conversational Recommender Systems