GroupNet

Login page
Profile page
Matches
Messaging

Inspiration

While platforms like Tinder are popular, they lack focus on the specific needs of academically-oriented university students who seek collaboration and study partnerships rather than dating interactions. However, our software matches people by more than just classes, allowing for strong friendships to be born!

What it does

GroupNet matches students based off of major, classes taken, the class you need help with, hobbies, and study methods to allow the user to find a study partner(s). A swipe left/right system of the best matches is implemented such that the user can choose the people they want to study with out of the most closely related people. Finally, they can message the people using the messaging system implemented into the account.

How we built it

GroupNet leverages sophistacated features and frameworks on the frontend and backend to run smoothly:

Frameworks (backend): Python, MongoDB, websraping, KNN (K-Nearest Neighbors Algorithm), LLM's, AI/ML, Flask

Frameworks (frontend): CSS, JS, HTML, Tailwind, React JS, Node JS, Express JS, REST API, Firebase, .JSON integration

We started by webscraping all the courses off the ASU website using Selenium and Beautiful Soup. Then, we augmented the scraped .JSON into classes sorted by major. We wrote a generate students script using LLM's to create a database of 5000 students with previous courses taken and needing help courses for accurate data. Then, we used a matching algorithm with endpoints so when the user entered the profile information on the frontend, the data would be recieved in the backend. We implemented KNN to match and returned the best matches back to the frontend, where the user can choose if they want to match with the person or not. Finally, we used firebase to implement a proper messaging system for users to message each other after they match!

Challenges we ran into

We struggled a lot with REST API integration as well as firebase API usage. However, we were able to overcome these challenges!

Accomplishments that we're proud of

We took on a really ambitious project for our skill level and we are really excited that it runs properly and we could actually turn a concept into working software!

What we learned

We learned version control really well and how to use databases at a higher level. We also were able to implement react, which is something we've struggled with previously.