An Improved Spam Detection Technique using Deep and Recurrent Neural Networks
M. S. Arunkumar, Deepa S, R Suguna, Thirunavukkarasu Kanimozhi, M Chandru, L. K. P. Vignesh · 2025
Short Message Service (SMS) is still a vital and indispensable form of communication in our everyday life, even in light of the swift advancement email and chat based textual message communication. Now-a-days majority of the companies discover that short textual based message takes more successful medium to connect with their audience than emails. The data indicates that the persons are viewing their mail inbox randomly and read the complete mails in 1:4 ratios. In another side, more than 80% of the short textual messages are viewed as well as replied within five minutes of being received. Spammers have taken notice of SMS due to its continued significance for mobile phone users. As the technologies advances emerges, security risks like Smishing (SMS-based phishing attempts), also considerably increase in recent years. The proposed work presents a novel hybrid deep learning model for SMS spam message identification. Here, we combine the potential of deep and recurrent neural network to create an elevated detection model. The proposed work is compared with few other deep learning techniques to get a comprehensive analysis of the proposed system. The experimentation done proves that the proposed RNN-DNN out performs the existing algorithms with an improved accuracy rate of 98.47%.