Library Book Classification System Using the Naïve Algorithm Bayes (Case Study: Politeknik Caltex Riau)

Abstract

Library is a room with a collection of books arranged in such a way, make it easy for readers to find. Based on interviews with Politeknik Caltex Riau Librarians, the collection of books was arranged according to the Dewey Decimal Classification rules. Librarians must estimate the appropriate subject for each book based on the DDC Guidebook manually. According to librarians, the process is not easy and takes a long time, due to the number of books that need to be classified and the difficulty of determining the right topic for the book. The implementation of Text Mining, Frequent Itemset Mining, and Naïve Bayes can solve the problem. Text Mining is needed to process title and synopsis of book, Frequent Itemset Mining to determine the frequent word, then Naïve Bayes algorithm is used to determine the classification number and corresponding subject. Based on UAT test results, can be concluded that system runs well and has appropriate to the expected. In Confusion Matrix testing, can be  obtained that the accuracy of the system was 80% and an error rate was 20%.