Book Recommendation Framework Using Collaborative Filtering with Hybrid User Overlap Modeling

This paper introduces a machine learning–based book recommendation system that enhances traditional collaborative filtering by incorporating a novel hybrid overlapping function. The system processes a dataset of roughly 270,000 users, 250,000 books, and 1.2 million ratings. By discarding data from users with fewer than 200 ratings and books with fewer than 50 ratings, the dataset was refined to 811 users and 706 books. An 810-dimensional matrix represents user ratings for each book, and cosine similarity is used to assess pairwise similarities between books. Additionally, our hybrid overlapping function calculates the ratio of shared user ratings (i.e., the intersection over union) to adjust the similarity scores. Extensive experiments, statistical analyses, and case studies demonstrate that this approach improves recommendation precision by ap- proximately 4.7%, reliably identifying the five most similar books to any given title and thereby enhancing user experience.