Depred

• 

  tweets

Inspiration

Seeing a drastic increase in suicide due to depression,we decided that this project will be very helpful in stopping kind of activities by finding out if a person is depressed before hand providing help.

What it does

It goes through the tweets database and applies lstm model and pretained embedding vectors to find out which of the tweets are felt to be depressive and then this can be flagged and maybe an ngo can contact them to help them out with the same.

How we built it

The speciality of our model is that we didnt use and pre existing data set rather we decided to make a twitter developer account and extracted tweets so that we get the latest data and factual data.Also we used a special tool called label-studio where we created a interface to label the tweets manually which helped us to do it much better than taking pre existing datasets.

Challenges we ran into

we ran into multiple challenges in choosing the model and while training,but we were able to complete it in the end.

Accomplishments that we're proud of

That our data is original and not been worked by anyone else before,plus it can help the society in a major way.

What we learned

we learned how to use the ibm tools which was good,also we learnt how to train different ml model .

What's next for Depred

Built With

• label-studio
• python