Intelligent book recommendation system using ML techniques

^1,2,3,4 Department of CSE, Vignan’s Foundation for Science, Technology and Research Vadlamudi, Guntur, Andhra Pradesh, India
⁵ Department of Computer Science and Engineering, Vignan’s Foundation for Science, Technology and Research, Vadlamudi, Guntur, Andhra Pradesh, India

¹ Email: bhagyaasri34@gmail.com

Abstract

The current research focuses on a recommendation system based on Decision Tree, Naive Bayes, Ridge Classifier, and Random Forest, using a new hybrid method combining Singular Value Decomposition (SVD) and K-Nearest Neighbors (KNN). The Decision Tree model reaches a good trade-off for precision, recall, and F1 metrics, acting as a benchmark. On the other hand, the hybrid model greatly surpasses the remaining ones in such a way that precision is as high as 89.35%, recall is 59.01%, and F1 is up to 71.30%, thus reinforcing the notion that it finds user preferences for recommendations more effectively. By combining collaborative filtering with similarity-based recommendations, the system can model user-item interaction and thus improve the quality of book recommendations. The above findings imply that the hybrid model holds good potential to deliver better recommendations in a more relevant and acceptable way, countering the implicit limitations of their single-algorithm counterparts and increasing user satisfaction.