PERBANDINGAN ALGORITMA NAÏVE BAYES DAN SUPPORT VECTOR MACHINE (SVM) DALAM KLASIFIKASI SMS SPAM BERBAHASA INDONESIA

Widyawati Widyawati (Universitas Banten Jaya)
Sutanto Sutanto (Universitas Banten Jaya)

Jurnal Ilmiah Sains dan Teknologi (SAINTEK)
Vol 3 No 2 (2019): Vol 3 No 2 Agustus 2019

SMS Spam (Short Message Service Spam) is an unwanted message, including advertisements and fraud. The direct impact of the spam is the inconvenience on the receiver side, therefore there is a need to have a spam screening process. One of the possible approach is to filter the spam by classifying the messages. In this paper, we compare classification performance by Naïve Bayes and Support Vector Machine (SVM). The process include preprocessing (tokenizing, stopwords removal, and stemming) from each training and testing dataset before process the information through Naïve Bayes and SVM. The results from the processing data of 1143 records (765 training and 378 testing) showed that Naïve Bayes performance surpassed SVM in term of recall (94%) and Precision (95%).