PPT slides: https://drive.google.com/file/d/1kzed-7Tl8qdhcbUxfMVeKcQHy_k6Ul0u/view?usp=sharing
Recording of presentation of project (including demo): https://drive.google.com/file/d/1OqalHPTDYaWkRXamURQvsC9rwGMYIiFE/view?usp=sharing
Detailed in case-study-gnod.
Summary: build a song recommendation engine using Billboard 100 songs and recommend users songs based on their input songs.
Detailed in presentation slides and recording.
spotify_song_prediction.ipynb - first part exports a dataset csv which are the top songs, this is to be done at regular intervals (e.g. weekly). Second part is for modelling and user interaction. Run last cell to enter user input.
Open spotify_song_prediction.ipynb and run all cells to start, where you will be prompted to enter user input. To enter a new user input song, run last cell again.
- Web scrapping
- K means clustering based on song features obtained through API.