This project aims to provide users song recommendations similar to a specific song they provide.
- Web Scraping: Lyrics were not included in the base dataset and are retrieved from Genius.
- Recommender System: The system employs the k-Nearest Neighbors (kNN) algorithm to find songs with attributes closest to a given song.
- Embeddings: NLP techniques such as embeddings are employed to incorporate song lyrics into the recommender system.
The project utilizes the 30000 Spotify Songs dataset available on Kaggle. This dataset of 30,000 songs includes song information such as name, artist, and many musical features like popularity, danceability, energy, and acousticness.