PERBANDINGAN KINERJA VARIASI NAÏVE BAYES MULTIVARIATE BERNOULLI DAN NAÏVE BAYES MULTINOMIAL DALAM PENGKLASIFIKASIAN DOKUMEN TEKS

Abstract

Classification of text documents with large amounts will be a job that requires a lot of time, effort and cost of having to read text documents and then categorize them manually, therefore the automatic classification of text documents is needed. The algorithm developed is K-Nearest-Neighbor (KNN), Naïve Bayes, Support Vector Machine (SVM), Decision Tree (DT), Neural Network (NN) and Maximum Entropy. The algorithm used as the object of research is a variation of the Naïve Bayes algorithm, the Naïve Bayes Multivariate Bernoulli and the Naïve Bayes Multinomial. This study discusses whether there are differences between the Algorithms. The Naïve Bayes Multivariate Bernoulli algorithm and the Naïve Bayes Multinomial can be seen from the value of the agreement and the speed of the process of classifying text documents, as well as more information about the process of processing requests that are getting more and more requested. While the highest value using the non-stemming Naïve Bayes Bernoulli method is 71.33%, and the fastest processing time is required using the non-stemming Naïve Bayes method which requires 0.12 seconds processing time.