This project is aimed to provide a music recommendation system that recommend songs based on people's personalities. Big five personality traits (five factor model (FFM) or OCEAN model) are useful when studying on psychological identities. Therefore, some academic researches and surveys on correlations between musical taste (mainly genres) and big five personality traits are heavily used in this project.

Spotify API integrated with recommendation system to maintain artists, albums, songs, and significant audio features to create a deep learning model which classifies songs regarding to their genres. Additionly, recommendation system accepts feedbacks from participants if recommended songs are not relevant in matter of musical taste, participants are able to send their genre preferences to improve model predictions.

Python, Django, PostgreSQL, TensorFlow, Keras, numpy, pandas, scikit-learn, MaterializeCSS, and many other technologies are used in development phase.

Developed during graduation thesis by Orçun Özdemir in Istanbul Technical University and licensed under a Creative Commons Attribution 4.0 International License.

Support to maintain, all donations are welcome!

BTC (Bitcoin): 153LVRoDEoTiR3qJEJZUGoadfAU3w23UHd
XMR (Monero): 44o3jRiPj6GCFMgUeNzqKzCu5DJrxFbXuKJo8zkWN41u8qsPKiN7aV2axms1QVCUmXRptuU55urJHSTz4SysGvPG5PwHzSp