Rekomendasi Jenis Buku Berdasarkan Minat Baca Menggunakan Naïve Bayes

Reading interest among junior high school students is important in supporting literacy and learning activities, yet students may have difficulty identifying reading materials that match their interests. This study aims to implement a Naïve Bayes algorithm in a web-based recommendation system for determining book types based on students' reading interests at SMP Negeri 3 Pagedongan. The research uses Research and Development (R&D) with the ADDIE model consisting of Analysis, Design, Development, Implementation, and Evaluation. Data were collected through observation, questionnaires, and literature study. Naïve Bayes processes questionnaire responses by calculating prior and likelihood probabilities for each book category and selecting the category with the highest posterior probability. The system was developed using PHP, MySQL, and Bootstrap and provides six book-category recommendations. White Box Testing was applied to the login, questionnaire, and Naïve Bayes modules, and all tested independent paths were valid. User evaluation involving 31 students produced an overall score of 4.05 on a five-point Likert scale, equivalent to about 81% and categorized as very good. The results indicate that the system can assist students in identifying book types aligned with their reading interests and provide supporting information for schools when considering library collection development.