This research introduces a novel Continuous Multi-Prompt Engineering approach that enables efficient mental disorder detection from text with minimal training data.

• Leverages large language models with injected medical knowledge to reduce dependency on extensive labeled datasets
• Implements a continuous prompt tuning method that outperforms traditional few-shot approaches
• Demonstrates effectiveness across multiple mental disorders with minimal training examples
• Creates a more accessible framework for mental health professionals without requiring deep ML expertise

This advancement significantly lowers the barrier to implementing AI-based mental health screening tools, potentially enabling earlier intervention and better patient outcomes in clinical settings.

Few-Shot Learning for Mental Disorder Detection: A Continuous Multi-Prompt Engineering Approach with Medical Knowledge Injection