Spam Email Detection with Data Size Reduction Using a Novel Hybrid Technique

Spam email is becoming a significant concern in network administrators' and social media users' lives. Spam email is unwanted and unpleasant email that has an impact on system security and may contain threats such as fishing, virus, or worm. It compels us to incur several costs, such as the use of network bandwidth, difficult calculations, high space and time complexity, and so on. Because spam emails are becoming more prevalent, and over ۸۵ percent of all received electronic mails are spam emails, recognizing these forms of emails are essential. The major aim of this study is the size reduction of data set without losing accuracy using data mining methods using a recommended novel hybrid model. The edge detection algorithm is used to find the edge of each cluster, the k-means algorithm is used to calculate the centroid of each cluster in this recommended method, and the outputs of these two algorithms are given to the Medoidshift algorithm to reduce the volume of the dataset. The experimental results show that the proposed method resulted in a significant reduction in data size without losing the accuracy. The results demonstrate the superiority of the proposed method over other methods.

Keywords:

Spam e-mail detection , Data mining approaches , Classifying e-mail , Filtering spams