Про методи фільтрації даних під час формування персоналізованих рекомендацій музичної комп’ютерної системи

Abstract

This work is devoted to the research of hybrid filtering methods for personalized music recommendation systems. The study examines content-based and collaborative filtering approaches, their key advantages and limitations. Popular streaming platforms such as Spotify and YouTube Music primarily rely on listening history, which leads to monotonous recommendations. The hybrid approach, combining both methods, allows to overcome these limitations and significantly improve the quality of personalized recommendations