Asperger Syndrome is a disorder that classifies under the autism spectrum. It is most commonly known to affect people’s social interaction and their ability to read other people’s emotions. This usually puts a strain on personal relationships, as affected people are typically more blunt in their expressions. The goal is to help them recognize stronger negative connotations in everyday chatting, which is reflecting in a more intense colour, giving real-time feedback from the Google NLP API.

What it does

With moody chat, the chatbox explicitly indicates the overall tone of a message using Google's Natural Language API, and reflects that tone in the form of chat bubble colour.

How I built it

We brainstormed ideas, and wanted to incorporate a Google ML API technology into our project with the goal of leveraging human connection for social good. We built a simple chatbox interface using Node.js and, and used the Express framework to connect everything together with HTML, CSS, and Javascript. The Google Natural Language Processing API is called in app.js and returns a sentiment score and magnitude.

Challenges I ran into

Setting up all the Google projects, client libraries, keys, etc. and getting authorization! We tried several different chatbox interfaces before we were able to run one smoothly. The majority of our team had no experience with web development/Node.js/Java!

Accomplishments that I'm proud of

Proud of getting the Google API to work with the chatbox messages!

What I learned

The majority of us had zero experience in web and back end development. We learned the basics of Node.js and "connecting" everything behind the scenes, including setting up the chatbox and connecting the Google API. Also, communication throughout the entire hackathon is very important to make sure everyone has the same vision of the final product :)

What's next for moody chat

Incorporation with Google Firestore for data storage (ex. users will have profiles), and web deployment (maybe using Google App Engine)

Built With

Share this project: