Skip to main content Skip to navigation

TheraGen: Therapy for Every Generation

Project Overview

The document explores the integration of generative AI in education, emphasizing its potential to enhance learning experiences and support mental health. A key application highlighted is TheraGen, an AI-powered mental health chatbot designed to provide accessible and personalized support to students and educators. Leveraging the LLaMA 2 7B model, TheraGen operates around the clock, utilizing a comprehensive dataset to engage users in meaningful conversations. The chatbot has garnered impressive user satisfaction, with 94% of participants reporting improvements in their mental well-being, illustrating its effectiveness as a complementary resource rather than a replacement for traditional therapy. The findings suggest that generative AI can play a significant role in addressing mental health challenges within educational settings, ultimately bridging gaps in service accessibility and enhancing overall student support systems.

Key Applications

TheraGen, an AI-powered mental health chatbot

Context: Accessible mental health support for individuals with mental health issues

Implementation: Utilized the LLaMA 2 7B model fine-tuned on a dataset of 1 million conversational entries, deployed on cloud infrastructure for scalability and availability

Outcomes: 94% user satisfaction, improved mental well-being for users, high accuracy in response generation (BLEU score of 0.67 and ROUGE score of 0.62), significant reduction in response time

Challenges: Not a replacement for professional therapy, ethical considerations regarding privacy and accuracy, maintaining user engagement and trust

Implementation Barriers

Ethical

Ensuring user privacy, transparency about AI capabilities, and addressing stigma associated with AI mental health support

Proposed Solutions: Implement strong data privacy measures, inform users of AI's role, and emphasize the complementary nature of AI support to professional therapy

Engagement and Trust

Users may be skeptical about the efficacy of an AI-based mental health tool compared to traditional therapy

Proposed Solutions: Provide consistent, high-quality interactions and demonstrate the value of AI over time

Accuracy and Reliability

Ensuring the accuracy and reliability of responses to maintain user trust

Proposed Solutions: Integrate advanced natural language understanding techniques and regularly update based on user feedback

Project Team

Kartikey Doshi

Researcher

Jimit Shah

Researcher

Narendra Shekokar

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Kartikey Doshi, Jimit Shah, Narendra Shekokar

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies