Komparasi Algoritma Naïve Bayes dan Support Vectors Machine pada Analisis Sentimen SMS HAM dan SPAM

Abstract

SMS is a form of communication in the form of messages sent using mobile phones between the designated numbers. SMS is now rarely used because many of the features that have changed are used by chat applications. However, the SMS feature was not removed for one thing, official messages from various applications for leveraging or other official information still use SMS as a sign that the phone number used is there. However, since 2011 there have been so many misuses of this function, so it is suspected that many frauds use SMS as a tool to influence victims. This sms category goes to SMS spam. Therefore, SMS needs to be classified so that users can find out that the SMS is included in the category of Spam or ham (the opposite of spam). Using 400 datasets taken from the UCI repository which is divided into two classes, namely spam and ham, we compare two classification methods, namely Naive Bayes and Support vector Machine in order to get SMS filtering correctly. And after the calculations are done, the accuracy is obtained in Naive Bayes, which is 90.00% Support Vector Machine 81.00%.