SMS Spam Classification and Through Recurrent Neural Network (LSTM) model

Short messaging service (SMS) spam is the unwanted messages delivered to the inbox of mobile devices from spammers. Service providers are worried about these spam messages as their clients get dissatisfied with services due to the spam data reaching on their mobile phone. There are most of the service providers has given facility Do Not Disturb (DND) activation for their clients to save them from most of the spam messages. Even though the spam messages are not controlled fully, the delivery of such messages are unstoppable. To overcome this issue extensive research has been done. Artificial intelligence made it possible with extensive learning model and accuracy of detection. This paper is proposed to classify short messages as spam or ham based on a deep learning model. In this paper, the spam detection through Recurrent Neural Network (RNN) model, in specific Long Short Term Memory (LSTM) model is used. The dataset used for this study is extracted from Grumbletext website and it has a total 425 short messages with ‘Ham’ and ‘spam’. The LSTM model classified the SMS dataset effectively with the learning model. Experimental study showed that the model has achieved an accuracy of 88.33% accuracy on SMS spam classification with the LSTM model.