Aashwasth(आश्वस्त)

We are working on building a decentralized app(Dapp) named Aashwasth(आश्वस्त) which helps to find out early signs of depression among people through their chat messages using a sentiment analysis model of emotion recognition.

🏆 Our hack was adjudged Winner (Sophomore Year) at the JUET Builds 2022 hackathon!

The problem it solves

We are the Generation of Stress. We are, more than seemingly ever before, dissatisfied by our realities. People are getting stressed out even about small things happening with them. This much of stress and anxiety eventually leads them to depression and even suicides.To help this at-risk population, we decided to create this platform!

Challenges we ran into

Learning and implementing Gun.js and sentiment analysis model, and integrating them in our code along with Node.js and JavaScript was quite a task, but we did that successfully. Through our model we have achieved an accuracy of 94%, which is good but we are trying to improve it. We also have to make the chat app much more user friendly.

Technologies we used

• Svelte
• JavaScript
• Node.js
• Gun.js
• TensorFlow
• Python
• HTML/CSS
• TFlite
• Bidirectional Long Short Term Memory (LSTM)