Inspiration

Our proposal aims to determine the risk of depression through analyzing text message to help psychologists with early detection of depression.

What it does

Analyses messages or tweets that are related to depression or show signs of depression to identify risks of depression among the users. Users with friends showing signs of depression can utilize our web app as well in order to justify if the people they are caring for are in a risk of getting depression and thus, able to take action early.

How we built it

Use of AWS services - comprehend, sagemaker

Challenges we ran into

Integrating front end with the models trained

Accomplishments that we're proud of

Successfully trained a custom entity recognizer and deployed it in entity recognition

What we learned

Learning how to use Amazon Comprehend to train the machine learning models. Learning how to use Amazon S3

What's next for Hashhs

Training more entity recognizers to detect signs of different mental illness

Share this project:

Updates