Enhancing Book Recommendation Systems: A Comparative Analysis of Collaborative Filtering Techniques

This study describes the creation and analysis of a book recommendation system that employs collaborative filtering techniques. It uses K-Nearest Neighbors (KNN) and matrix factorization techniques to make individualized book recommendations. The system is built on a dataset of user ratings and book metadata, relying on measures such as cosine similarity to calculate user and item associations. Experimental results show that, while KNN-based methods are computationally efficient, matrix factorization techniques provide better prediction accuracy. This work emphasizes the trade-offs between accuracy and efficiency, using hyperparameters to properly balance these elements.