Building a virtual psychological counselor by integrating EEG emotion detection with large-scale NLP models

We present a virtual counselor pipeline combining EEG emotion detection (feature extraction, elicitation protocols, classifiers) with LLM-based therapeutic prompting. The paper outlines two orchestration paradigms—direct suggestion and prompt method—and discusses limitations and future directions towards real-time assistive systems.

• EEG modules: emotion models, datasets, elicitation, feature engineering, classifiers.
• LLM modules: dialogue control, safety scaffolding, and task prompting.

Accessible abstract and PDF mirror available online.

Recommended citation: Lin P, Yang L. (2024). Building a virtual psychological counselor by integrating EEG emotion detection with large-scale NLP models. Proceedings, 12924:3013169.
Download Paper