LSTM Based Short Message Service (SMS) Modeling for Spam Classification

The Short Message Service (SMS) has widely extended in the modern methods of communication technology. The classification of the spam message is an interesting and prominent issue. Classifying availability of spam in SMS is a challenging task, a plenty of research has been carried out in this direction employing Machine Learning techniques such as Naive Bayes (NB), Random Forest (RF), and Support Vector Machine (SVM) for Spam Classification. Although these methods have shown adequate performance, but are not efficient enough in terms of spam classification. Hence, a rigorous study is needed to find a more accurate and robust method. To address this, we proposed a novel method Long Short-Term Memory (LSTMs), which is an advanced structure of Recurrent Neural Network (RNN) that has gating mechanism including memory cells. Additionally, Word2Vec tool has been employed in this study, which converts simplified text into representation of words in a vector space. To evaluate the effectiveness of our method, SMS datasets have been used which are freely available. Experimental results prove that proposed method outperformed state-of-the-art Machine Learning methods like Random Forest (RF), SVM, kNN (k Nearest Neighbor), Decision Tree, and providing 97.5 percent accuracy.