SMS Spam Detection Using Machine Learning
Suparna Das Gupta, Soumyabrata Saha, Suman Kumar Das
Abstract
Abstract In the modern world where digitization is everywhere, SMS has become one of the most vital forms of communications, unlike other chatting-based messaging systems like Facebook, WhatsApp etc, SMS does not require active inter n et conne ct ion a t all. As we all know that Hackers / Spammer tries to intrude in Mobile Computing Device, and SMS support for mobile devices had become vulnerable, as attacker tries to intrude to the system by sending unwanted link, with which on clicking those link the attacker can gain remote access over the mobile computing device. So, to identify those messages Authors have developed a system which will identify such malicious messages and will identify whether or not the message is SPAM or HAM (malicious or not malicious). Authors have created a dictionary using the TF-IDF Vectorizer algorithm, which will include all the features of words a SPAM SMS possess, based on content of message and referring to this dictionary the system will be classifying the SMS as spam or ham.