Аналіз взаємодії алгоритмів content-based та collaborative filtering у рекомендаційних комп’ютерних системах

Abstract

This work analyzes the principles of content-based and collaborative filtering algorithms in modern recommendation systems. The advantages and disadvantages of both approaches are considered. Content-based recommendation systems typically demonstrate higher effectiveness when operating with a limited amount of user data, whereas collaborative filtering requires a sufficiently large history of user interactions to achieve reliable results. Propose using both recommendation methods, in which content-based filtering is applied at the initial stage of interaction with the user, and collaborative filtering is activated after a sufficient amount of data about the user’s preferences has been accumulated