[Felip Samitier Jerez]
[DA, Barcelona & 2021]
⦁ This project is based on a song given by the user. If this song is among the 100 "Top" songs, it returns a recommendation from this list, otherwise, it extracts the characteristics of the song, classifies it in a group of "Clusters", returns a song from the same "Cluster" and with similar characteristics.
As an extra, we have added a context to the music recommender, i.e. the user can choose for which situation he/she is going to listen to that song. Recommending a song that fits the context and the characteristics of the song entered by the user.
(Image Workflow Implemented in slides)
⦁ User introduce one song.
⦁ Check if this song is in the "top 100".
⦁ If it is in the "top 100" it returns a song from this same group of songs.
⦁ If it is not in the "top 100", search the song on spotify, pull out the characteristics of the song.
- It recommends a song with the same characteristics.
⦁ It asks you for your context.
- Recommends a song depending on the context.
⦁ My folder consist of:
- One folder that contains all data
- Other folder that contains images for slides
- Other folder that contains the code for get the features
- Final Code: Final Recommendator
- Tweo files that are the pickles
- And two files (ignoreme and readme)
⦁ Slides