himalisaini / Mental-Health-Support-Bot

Developed as a part of Hackathon for Mercor

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Mental-Health-Support-Bot

Developed as a part of Hackathon for Mercor The main logic is present in the textbase/main.py file

Mental Health Support Bot

Overview

Welcome to the Mental Health Support Bot project! This is an empathetic AI chatbot, trained using OpenAI's GPT-3.5 model. The chatbot is designed to offer mental health support to users, aiming to promote positivity and reassurance through meaningful conversation.

Features

  • Sentiment Analysis: The chatbot uses sentiment analysis to comprehend the mood of the user's input. This analysis helps in formulating appropriate responses.
  • Dynamic Prompts: Based on the sentiment and conversation history, the bot selects from a variety of prompts (reassuring, encouraging, solution-oriented, wrap-up) to guide the conversation in a sensitive and understanding manner.
  • External APIs: The bot uses external APIs to fetch positive quotes and jokes, adding a unique aspect to its responses when the situation is suitable.
  • Logging: The chatbot maintains a log of the prompts used and sentiment analysis of each message in the 'prompt_history.txt' file.
  • Emergency Handling: The bot is designed to recognize severe situations based on its own responses and specific user inputs. In these cases, it advises the user to seek help from mental health professionals.

Purpose

The Mental Health Support Bot provides a safe and non-judgmental platform for people to express their feelings. It offers encouragement and positivity, thereby supporting mental well-being. It is important to note that while the bot can offer support, it is not a replacement for professional help.

Note

While this chatbot is designed to support users in distress, it is crucial to remember that in severe situations, professional help should always be sought.

About

Developed as a part of Hackathon for Mercor


Languages

Language:Python 65.0%Language:TypeScript 26.6%Language:CSS 4.0%Language:JavaScript 2.9%Language:HTML 1.5%