Relevancy Test

• 

  Home Screen: file upload page

• 

  Result Screen: shows participants

Inspiration

One of our member's parent is a 4th grade teacher and one thing she said she has trouble with when teaching online is that she has no idea how engaged students are. Sometimes students chat about unrelevant topics, like playing Roblox or Among Us, or say something unrelated, like "my mic isn't working". So, we built this to help teachers see how engaged their students are.

What it does

Logging in via google authentication, users can upload the chat file and the video transcription file into the web application and a machine learning model will run to determine how relevant the phrases are. From there, a list of participants are shown with an option to see one participant's contribution over time.

How I built it

The google authentication is built using firebase. The frontend is built using Angular and Angular Material. Finally, the machine learning model is built using python and pytorch.

Challenges I ran into

Communication was a problem as we were located in different states. Our solution was to have video call on at the side so that we could easily talk with each other as we worked.

Accomplishments that I'm proud of

We are proud of being able to make the different parts needed. In addition, we are proud of the work we were able to accomplish in the time that we had.

What I learned

Since some of our team was new to machine learning, we got to learn machine learning.

What's next for Relevancy Test

The machine learning model will definitely need to be trained some more. And, we hope that one day we can make it recognize topics.

Built With

• angular.js
• firebase
• python
• pytorch