Abstract:

The rise of network technology has led to notable advancements in music recommendation systems, making online music platforms the preferred choice for people to listen to their favorite songs. Nevertheless, these systems encounter a range of obstacles, including difficulties in data storage, suboptimal computational efficiency, issues with cold start, and sparse data due to the vast amount of information involved [1]. To tackle these challenges, the objective of this study is to create and establish a music recommendation system that effectively addresses these kinds of concerns. The approach is implemented using K Nearest Neighbor Classification and Random Forest Classifier. Since this paper proposes another approach for the recommendation, the other approach utilises the Spotify API calls. It aims at the implementation using the various machine learning algorithms which includes K Nearest Neighbor Classification, Decision Tree Classifier and the Random Forest Classifier. This model recommends the music to the user by taking certain factors into account like genre, year range and music features like acousticness, danceability, energy, tempo, valence. The Spotify API is another approach that will recommend the songs based on the previous listening history, preferences, likes of a particular user which comes under the content based filtering.

Description:

Content based filtering, K-Nearest Neighbor, Machine Learning, Random Forest Classifier, Recommendations, Spotify API