Inspiration

We all got together as a team and decided that we wanted to make something surrounding music and insight through data, and from that came the joke that there is a Drake song for every mood. From here, we decided to create a web application that can determine a sentiment analysis score from a word or phrase inputed, and provide you with a song that fits your mood perfectly!

What it does

Mood Matcher takes in a phrase or word that you provide about how you are currently feeling! Example:

I'm feeling sad, my girlfriend broke up with me

Marvin's Room by Drake is the best song for you right now!

How we built it

We built this using React Javascript on the front end and Python + Flask on the backend. Totally a weird stack but it worked! We wanted to play on the strengths of all of our team members and it worked out well.

Challenges we ran into

We had a LOT of troubles and conflicts when trying to find sets of data and sets of lyrics to songs that we can run sentiment analysis onto. While we work out the kinks on a much broader song library, we will be rolling out the demo with a single artist, Kendrick Lamar, and then expanding the library of songs from there.

Accomplishments that we're proud of

Everyone working on the backend of this project got to work with and create sentiment analysis for the first time! WE used various APIs including the Genius API, Spotify API, and some Textblob for sentiment analysis.

Matthew, working on the frontend, made a comprehensive front end in react, after previously not knowing how to work with react from start to finish.

What we learned

This weekend, we learned as a team about how to implement sentiment analysis

What's next for Mood Matcher

Share this project:
×

Updates