An Empirical study of Short Messaging Service - Smishing Detection using Machine Learning Algorithms

Phishing attack with SMS basically, smishing
occurs when a SMS message which looks legitimate makes
the common make believe that it’s a the genuine message
to open and click the links which may be malicious. Online
surveys prove that phishing attack as the leading attack for
web content which targets the financial concerns. Smishing

is a way to detect the SMS phishing techniques. Now-a-
days, Smishing is a matter of concern which creates havoc

among the common people. The individuals are not aware
of the phishing techniques and they fall prey to the fraudster
messages and shell out a huge sum from their savings. Also,
the awareness about the same is given by the government
and many social media, to the people, but then too these
kinds of frauds are still happening. This paper shows the
empirical study on the detection of smishing attacks using
machine learning algorithms like SVM, Random forest and
their ensemble models. The most appealing result is
achieved by the ensemble model with Support Vector
Machine and Random forest using voting classifier with
accuracy of 99.01% implementing with the SMS UCI
machine learning repository dataset.