Music Recommender System Using K-Nearest Neighbor and Particle Swarm Optimization

Randika Dwi Maulana Rasyid (Telkom University)
ZK Abdurahman Baizal (Computational Science, Faculty of Informatics, Telkom University)

Indonesian Journal on Computing (Indo-JC)
Vol. 7 No. 2 (2022): August, 2022

In this day, users can listen to music anytime digitally and access them through the already available applications. A music recommender system is needed to help users choose music according to their interests and find music to listen to. K-Nearest Neighbor (KNN) is a popular method used in Collaborative Filtering (CF). In many studies, CF with the KNN method has been widely used, but it does not provide good performance. Thus, in this study, we use KNN, which will be optimized using Particle Swarm Optimization (PSO), which can improve the performance of the results obtained against the method used. System testing is done by comparing the performance of the KNN algorithm with the optimization results of KNN-PSO with several variables being observed, including the Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) values. The results of these recommender will predict the rating value where the KNN method gives MSE 4.48 and RMSE 2.54 while the KNN-PSO method gives MSE 1.70 and RMSE 1.30.