SMS Spam Detection Using Deep Learning Techniques: A Comparative Analysis of DNN Vs LSTM Vs Bi-LSTM

SMS spam detection is crucial for ensuring text messaging systems' security and usability. Spam messages, also known as unwanted text messages, can be a nuisance and be used for phishing scams and other malicious activities. In this research, we analyze various SMS spam detection methods and their efficiency in recognizing spam messages. Three machine learning models-LSTM, a DNN, and a Bi-LSTM model have been used to examine the data. We found that our proposed methods are highly effective at detecting spam messages with high accuracy through our experiments on a dataset of SMS text messages. Of the three models, the DNN has produced the highest accuracy of 95.6522%, followed by Bi-LSTM with an accuracy of 93.9799% and LSTM with an accuracy of 92.9766%.