Improved Email Spam Detection Using Integrated Approach of BiGRU-A-LSTM Approach

In today's world, email is used by nearly everyone. It has been called the most efficient, widely used, and low-cost method of communication currently available. However, the widespread use of email has been hampered by the proliferation of unwanted and potentially fraudulent messages that must be quickly identified and isolated using a spam detection system. Spam identification is crucial for the safety of email users and for avoiding the many harmful applications to which email has been put recently. There is a pressing need for improved spam detection techniques due to the adaptive nature of unsolicited emails that are sent through mailing programmes, which has severely reduced and sometimes rendered the efficiency of existing spam detection methods. Despite the fact that several algorithms for detecting spam have been presented and tested, the reported accuracy indicates that further research is needed in this area. Preprocessing, feature selection, and model training are all used in the suggested method. Preprocessing makes use of tools like HTML tag removal, stop word removal, Tokenization, and word frequency. Feature selection is performed using TF-IDF. CNN-A-BiLSTM is utilized for model training. The proposed method has superior results to those of LSTM and BiLSTM, two well-established alternatives.