LLM 模拟来访者与真人角色扮演在在线咨询训练中的咨询模式对比分析
Comparing Human Role-Players and LLM-Simulated Clients in Online Counselling Training: An Analysis of Counselling Patterns
AI 导读
一项研究比较了 Mistral 的 Mixtral 8x7b（4-bit 量化）模拟的来访者与真人扮演的来访者在在线咨询培训角色扮演中的表现，在句子层面用咨询沟通模式系统分析新手咨询师与两类来访者的对话。结果显示两种场景下咨询模式的应用既有相似也有差异，表明 LLM 角色扮演有潜力提升咨询能力，但虚拟来访者模拟仍需改进。该短文发表于 EDM 2025。
正文 · 原文
Abstract:This study investigates the capabilities of Large Language Models to simulate counselling clients in educational role-plays in comparison to human role-players. Initially, we recorded role-playing sessions, where novice counsellors interacted with human peers acting as clients, followed by role-plays between humans and clients simulated by Mistrals Mixtral 8x7b using 4-bit quantization. These interactions were analysed with a counselling communication pattern system at sentence level. We investigated two key questions: (1) to what extent LLM-generated responses replicate authentic conversational dynamics and (2) whether counsellors' communication behaviour differs when interacting with human versus LLM-simulated clients. The findings highlight both similarities and differences in the application of counselling patterns across scenarios, showing the potential of LLM-based role-playing exercises to enhance counselling competencies and to identify areas for further refinement in virtual client simulations.
| Comments: | Published at EDM 2025 (18th International Conference on Educational Data Mining), short paper. 7 pages |
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.40563 [cs.AI] |
| (or arXiv:2610.40563v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.40563 arXiv-issued DOI via DataCite |
|
| Journal reference: | Proceedings of the 18th International Conference on Educational Data Mining (EDM 2025), short papers |
| Related DOI: | https://doi.org/10.5281/zenodo.15870260
DOI(s) linking to related resources |
Submission history
From: Eric Rudolph [view email]
[v1]
Thu, 29 Oct 2026 17:41:59 UTC (9,039 KB)
来源：arXiv cs.AI · arxiv.org