International Journal of Scientific Research and Engineering Development

This paper presents a comprehensive Book Recommendation System (BRS) built as part of an MSc Computer Science research project. The system addresses the problem of information overload in digital libraries and e-commerce book platforms by employing three complementary recommendation strategies: Content-Based Filtering (CBF), Collaborative Filtering (CF) via Singular Value Decomposition (SVD), and a weighted Hybrid approach combining both. The CBF module utilizes TF-IDF vectorization over book metadata (title, author, publisher) followed by cosine similarity scoring, while the CF module decomposes a sparse user-item rating matrix into latent factors. The Hybrid model blends both output scores using a configurable weighting scheme. The system exposes a RESTful API built with Flask and includes a responsive web frontend. Evaluation on a synthetic dataset of 500 books, 200 users, and 5,000 ratings yields Precision@10 of 0.0821, Recall@10 of 0.1134, NDCG@10 of 0.1056, and MAP of 0.0978 for the Hybrid model, outperforming both individual methods. The paper discusses architecture, algorithms, experimental results, and future directions including deep learning-based embeddings.

Shifa Bilal Tamboli, Simeen Phiroj Mulani, Arman Tajuddin Shiakh,"Book Recommendation System Using Hybrid Collaborative and Content-Based Filtering Approaches" International Journal of Scientific Research and Engineering Development, V9(2): Page(2279-2283) Mar-Apr 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.