Malay SMS spam detection framework using naïve bayes technique

Abstract

Short Message Service (SMS) Spam is one form of mobile device attack that can affect mobile user’s security and privacy. This is because such attack applies social engineering method to trick the user for information gathering. This study proposed an SMS Spam detection framework specifically for Malay language by using Naïve Bayes. There are several solutions to detect SMS Spam, but machine learning is one of the most effective technique to detect spam attack. In addition, the existing detection framework using machine learning technique is not effective for Malay language SMS. This is because the features used are not based on Malay language to detect the SMS content as Spam or not Spam. This framework consists of several processes such as Data Collection, Pre-processing, three types of Features Selection, Classification and Detection. Based on the result, it shows that the classification derives acceptable accuracy which is over 90%.