Skip to content

nolita26/Stress-Not-Application

 
 

Repository files navigation

AI-Based Social Media Analysis for Mental Health Evaluation

A system which provides assistance to the user by analyzing his state of mind by his social media contents and further recommending health care experts to get through their declining mental state. Analysis will be illustrated in a graphical format, which will keep updating on a daily or weekly basis while fetching live data.

To analyze the tweets and classify whether it is positive, negative or neutral and assign a score to each tweet and further detect the level of stress/depression.

  1. User’s live Twitter data User login to his Twitter Account and further analyzes his metal health.

  2. Weekly Statistics and Analysis Graph of user’s mental health progressed over a period of time

  3. Daily usage of Twitter The amount of time he spent on Twitter on a daily basis displayed statistically

  4. Alert the user if the stress level exceeds a certain limit

  5. Counting the tweets on the basis of 3 categories: +ve -ve neutral

Technology Stack: Python, NLP, Flask, Cloud Firestore, Twitter OAuth 2.0, Google Maps API, Twilio API

About

Developed an application that detects user’s mental state through social media by live data fetch from REST API (OAuth 2.0), utilizes CNN model, with Flask, alerts user using Twilio, and medical chatbot by DialogFlow and handles collection of live data through Cloud Firestore with 89.28% resulted test accuracy.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Dart 65.7%
  • Python 17.1%
  • Jupyter Notebook 16.9%
  • Other 0.3%