twentyOneSongs

Song recommender for songs by the Twenty One Pilots using structured song data.

What it is

A very simple song recommender and browser for songs by the Twenty One Pilots. It can plug in and out multiple technologies of your liking once they're implemented with one click.

Status

Server backends:

• [DONE] Python/Django
• [TODO] Python/Flask
• [TODO] Ruby on Rails
• [TODO] Node.js

Frontends:

• [DONE] Static HTML
• [IN PROGRESS] React.js
• [IN PROGRESS] AngularJS
• [TODO] Meteor.js
• [TODO] Ember.js
• [TODO] Backbone.js

APIs:

• [DONE] Python/Django handmade
• [TODO] Python/Django REST
• [TODO] Node.js
• [TODO] Firebase

Recommendation

Using good old business intelligence knowledge, the idea is that if we structure the songs using more dimensions than usual, we can provide a higher level of recommendation. I decided to add moods, instruments, pacing and other elements so we have more data dimensions for recommendation, which is viable for a small dataset of ~30 songs, and make an elite recommendation system for any song by the TØP.

Data

I also open-sourced the data I manually input - both the actual database scaffolded by Django and JSON exports are available for you to use in any pet projects of your liking.