Inspiration
Anthony Fantano is well known for his reviews and his unique taste in music. We wanted to try and replicate his taste through machine learning while also learning a new skill.
What it does
It takes a sample of past album reviews that Fantano has published and trains a model to predict scores using an album's "audio-features" as provided by the spotify API.
How we built it
We made a lot of CSV files to store extrapolated data from the spotify API and then used a RandomForestRegressor model to somewhat accurately predict Fantano's album rating.
Challenges we ran into
We didn't know what we should extrapolate We are new to machine learning so learning the processes needed was a challenge indeed The ratings our model creates is not as dynamic as we would have liked (gravitating towards a 5 at all times)
Accomplishments that we're proud of
Took our first steps in machine learning Learned a lot about navigating the spotify API somewhat effectively CREATED AN ML MODEL
What's next for fantanoBot
Implement more relevant data points to correlate towards the final score (effectively make the score more dynamic)
- Genre bias
- Artist bias (perhaps scan twitter interactions lol)
Built With
- python
- sklearn
- spotipy
Log in or sign up for Devpost to join the conversation.