Дослідження використання алгоритмів для персоналізованих рекомендацій настільних ігор

Abstract

This work is devoted to the analysis of recommendation algorІТhms for board games based on the content-based approach. The study explores various recommendation methods, including vector-based similarІТy calculations and hybrid techniques to improve accuracy. The problem of cold start and data normalization is addressed, ensuring better personalization for users. AddІТionally, the effectiveness of the proposed system is assessed using precision, recall, and F1-score metrics. The research highlights the importance of combining multiple filtering techniques for enhancing recommendation relevance.