Spam email is one of the most dangerous threats in the world. These spam mails are responsible for huge financial loses. The main intention of these spam mail is to stole the personal information from the individual like bank account numbers, cvv etc.., Now a days these spam mails are increasing day by day. Therefore, an effective technology is needed to identify spam mails precisely. In this paper, we mainly concentrated on Recurrent Neural Network model where the output of previous state acts as input to the current state for identifying spam mails. To find the effectiveness of RNN we used an unbalanced dataset which contains spam and normal emails. We also did a comparative study on RNN over two different algorithms like KNN and logistics and proved that RNN gives more accuracy rate in detecting spam emails.
Volume 12 | Issue 2
Pages: 845-853
DOI: 10.5373/JARDCS/V12I2/S20201104