UKPLab/Graph2Counsel

Graph2Counsel是一个基于客户心理图（CPGs）生成的合成心理咨询对话数据集。该数据集提供了从真实心理咨询会话中提取的CPGs、基于这些CPGs生成的多样化客户档案和对话，以及与每个CPG对应的咨询师策略。与现有的合成心理咨询数据集相比，Graph2Counsel在特异性、咨询师能力、真实性、对话流畅性和安全性方面表现出更高的专家评估性能。数据集结构包括患者标识、会话标识、变体索引、CPG、客户档案、咨询师策略和对话内容等字段。该数据集适用于微调大型语言模型（LLMs）以生成心理健康咨询对话，训练模型以基于结构化临床知识（CPGs）生成治疗响应，以及研究客户心理图、咨询师策略和对话质量之间的对齐。

Graph2Counsel is a synthetic counseling session dataset generated from Client Psychological Graphs (CPGs). The dataset provides the CPG, generated diverse client profiles and dialogues from this CPG as well as counselor strategies corresponding to each CPG collected from the real counseling sessions. This datasets shows improved expert evaluation performance over existing synthetic counseling datasets on specificity, counselor competence, authenticity, conversational flow, and safety. Each record contains fields such as patient identifier, session identifier, variation index, CPG, client profile, counselor strategies, and dialogue content. This dataset is intended for fine-tuning LLMs for mental health counseling dialogue generation, training models to ground therapeutic responses in structured clinical knowledge (CPGs), and studying the alignment between client psychological graphs, counselor strategy, and dialogue quality.