We wanted to help Aggies form the most cohesive and meaningful sentences while still incorporating the words we love.

What it does

Allows user to create sentence with buttons for common Aggie words. Then, graphics are made displaying the words to help improve sentences in future.

How we built it

We built the email builder using React.js, Bootstrap and the analyzer using Streamlit. The Streamlit code was all written in python.

Challenges we ran into

Andrew: Using React to update DOM without refreshing page Rishi: Saving the model in a Linux VM was different from saving it in Windows, so I had to rerun my NLP classifier again locally. Deploying to the cloud was also a lot of learning.

Accomplishments that we're proud of

A service which lets you construct emails with minimal input required. It then analyzes the email using NLP (Natural language processing)

What we learned

Andrew: learnt more about web development and integration in React.js and Firebase Rishi: learnt how to do machine learning in the cloud and present it in Streamlit. I also learnt more about Google Cloud Platform and ways the app may be deployed online.

What's next for The Aggie Jargon Challenge

Integrate the two parts of the application. Make the NLP model better by using more realistic data.

Built With

Share this project: