期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：24
页码：6788
出版社：Journal of Theoretical and Applied
摘要：Nowadays a monumental amount of spam and junk email clutter email inboxes and storage facilities. Spam email has a significant negative impact on individuals and organizations alike, and is a serious waste of resources, time and effort. The task of filtering spam or junk e-mail is complex and very difficult to solve. Hence, learning-based filtering is considered an important method for detecting spam emails as the filtering technique requires training to epitomize the knowledge that can be used for detecting the spam. Thus, Artificial Neural Networks are being relied on to create a learning based filter. In this article, we particularly propose the Feedforward Neural Network (FFNN) for identification of e-mail spam; the weights and biases of this network model are set to optimum using a new modified bat algorithm (EBAT). Experiments and results based mainly on two datasets (SPAMBASE and UK-2011 WEBSPAM datasets) show that the developed FFNN model trained by EBAT achieves high generalization performance compared to other optimization methods.
