SMS SPAM DETECTION USING DEEP LEARNING

Abstract— Short Message Service (SMS) is one of the mobile messaging apps that enables quick and simple and reasonably priced communication. The main problem in this is creating undesired messages for the purpose of advertising or harassment and sending these messages via SMS. Service. Unwanted short message spam has been detected using a variety of techniques, many of which are machine learning-based. In order to distinguish between SMS's legitimate brief messages (known as ham) and unwelcome text messages (known as spam), neural networks have been used. Recurrent Neural Network (RNN) hasn't yet been applied in this problem, as far as we know. Although we used predetermined sequence lengths in this study, we offered a novel approach that makes use of RNN to distinguish between ham and spam. The accuracy of the suggested method, which was 98.11, shows a significant improvement.