Our app view
Build a recommendation tool that is more reliable and diverse than other known music streaming platforms.
What it does
Provide an ideal collection of classical pieces to the users by generating a recommendation based on users' interest in composers, era, instruments, and etc.
How we built it
We used python for data scrapping, pure HTML, CSS, and JS for designing our front end, and Google Cloud App Engine for hosting our app.
Challenges we ran into
Building a custom API that scrapes the collection of classical pieces from the IMSLP Petrucci Music Library. Also, we had a problem hosting a website with a custom domain on the Google Cloud App Engine, namely hosting a dynamic web page powered by python flask server was the main cause of our problem.
Accomplishments that we're proud of
We had implementation ideas that can theoretically surpass the known song recommendation algorithms' capabilities.
What we learned
We learned about data scrapping using python. Fundamental components of Google Cloud App Engine.
Implementation approaches for randomizing samplings based on the user's interest. How to use .git & avoid merge conflicts.
What's next for Fermata - Classical Music Recommender
Relocate to a stable web hosting framework; allow more randomness in samplings for better-optimized collection. Add a feature that allows users to populate the initial classical music samplings.