Racacoonie

• 

  Let 'em cook!!!

Inspiration

We enjoy cooking and would like to create/share recipes from different cultures. Moreover, we also want to have a user recommendation algorithm that recommends users recipes that they would potentially like based on other users who use our platform.

What it does

A user recommendation system

How we built it

Front-end: We used react, material-ui and tailwindcss to create our frontend. Used Firestore for Oauth and web token creation.

Backend: We used flask to host of api endpoints. We used cockroach DB and firestore as the database.

Algorithm: We Implemented a collaborative filtering recommender system from scratch using cosine similarity to figure out the most similar users to yourself and recommend what recipes they liked. The idea is that similar users have similar interests (recipes)!!

Challenges we ran into

• Dependency issues with npm
• Integration issues with different components
• Data pre-processing issues
• Database issues

Accomplishments that we're proud of

Fully functional frontend that has recipe filtering, map Functional user recommendation system

What we learned

Importance of data processing. Importance of differences in package versions and data dependencies. Scoping projects and how to measure the scope.

What's next for Racacoonie

Improvement in the UI Integrating collaborated filtering Idk, become a chef or something

Built With

• cockroach-db
• firebase
• firestore
• flask
• javascript
• kaggle
• material-ui
• npm
• numpy
• pandas
• python
• react
• tailwindcss