Trading Fours is your own personalized Spotify recommendation engine.
Important Update: (I sadly took Trading Fours down on 12/03/2024, as it was rendered no longer functional when Spotify deprecated some core API endpoints I was using to get track features: (Audio Features & Related Artists)
- Search for a playlist or song, or browse your own library of saved playlists directly on the site, and Trading Fours will recommend alike tracks based on genre & track attributes
- Its thinking is influenced by what you have been listening to as of late, and will use that data to tailor the songs it recommends to you
- Behind the scenes, user and song info is being safely handled and stored using MySQL, and the bulk of the recommendations is being done via Python, where playlist / track data is being processed and computed in a ~45 dimensional space against a dataset of over 200,000+ songs (and growing!) to provide the best recommendations
Video:
Email: wangstanley910@gmail.com
University Email: stanley.wang@mail.mcgill.ca
LinkedIn: https://www.linkedin.com/in/stanley-utf8/
The initial dataframe of ~200,000 songs was collected and altered from the following:
However, the dataframe is continually updated by playlists and tracks that users enter!
Logo Notes by: rawpixel.com
Thank you, to anyone who checked out this project! It took me the better half of a year from start to finish, and I am incredibly proud of it.