Toward Real-World Chinese Psychological Support Dialogues: CPsDD Dataset and a Co-Evolving Multi-Agent System

2025-07-10

AI Summary

To address the scarcity of high-quality Chinese psychological support data and poor adaptability to real-world scenarios, this paper proposes a preset-path-guided dialogue generation and refinement framework. We construct CPsDD—the first large-scale, high-quality Chinese psychological support dialogue dataset—comprising 68K dialogues spanning 13 user demographics and 16 problem categories. Furthermore, we design CADSS, a collaboratively evolving multi-agent system integrating user profiling, strategy planning, and empathetic response generation. Technically, CADSS combines large language model fine-tuning, dialogue state tracking, and a hybrid human–automated evaluation mechanism. Experiments demonstrate that CADSS achieves state-of-the-art performance on both strategy prediction and emotional support tasks across CPsDD and ESConv benchmarks, significantly enhancing personalization and structural coherence in psychological support delivery.

Technology Category

• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Multiagent Systems: Adversarial Agents
• Natural Language Processing: Conversational AI/Dialog Systems

Application Category

• User Modeling, Personalization and Recommendation: Psychology-informed user models and recommender systems
• Search and Retrieval-Augmented AI: Assisted, interactive, and conversational search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The growing need for psychological support due to increasing pressures has exposed the scarcity of relevant datasets, particularly in non-English languages. To address this, we propose a framework that leverages limited real-world data and expert knowledge to fine-tune two large language models: Dialog Generator and Dialog Modifier. The Generator creates large-scale psychological counseling dialogues based on predefined paths, which guide system response strategies and user interactions, forming the basis for effective support. The Modifier refines these dialogues to align with real-world data quality. Through both automated and manual review, we construct the Chinese Psychological support Dialogue Dataset (CPsDD), containing 68K dialogues across 13 groups, 16 psychological problems, 13 causes, and 12 support focuses. Additionally, we introduce the Comprehensive Agent Dialogue Support System (CADSS), where a Profiler analyzes user characteristics, a Summarizer condenses dialogue history, a Planner selects strategies, and a Supporter generates empathetic responses. The experimental results of the Strategy Prediction and Emotional Support Conversation (ESC) tasks demonstrate that CADSS achieves state-of-the-art performance on both CPsDD and ESConv datasets.

Problem

• Lack of non-English psychological support dialogue datasets
• Need for scalable real-world counseling dialogue generation
• Developing multi-agent system for personalized emotional support

Innovation

• Fine-tune LLMs for psychological dialogue generation
• Co-evolving multi-agent system for support
• Automated and manual dialogue refinement