Its main theme is the neuro-fuzzy inference engine which has been applied successfully in a wide range of research. Initially, a normalization task is performed to remove any deceptive character from the text body. The preparation stage then examined the content of well-known datasets of spam emails to identify all the phrases that solely identify spam email. Our experiments show that IHASS could achieve very good accuracy level and works stably well. | International Journal of Computer Networks and Communications Security VOL. 5, NO. 6, JUNE 2017, 115–125 Available online at: E-ISSN 2308-9830 (Online) / ISSN 2410-0595 (Print) IHASS: Intelligent and Hybrid Anti-Spam System Shubair Abdulla1 and Altyeb Altaher2 1 College of Education, Instructional and Learning Technology Depart, Sultan Qaboos University, Oman 2 Faculty of Computing and Information Technology in Rabigh, King Abdulaziz University, Saudi Arabia 1 shubair@, 2aaataha@ ABSTRACT Spam emails are increasing continually and represent real threat to internet community. Email spammers use advanced software techniques to send millions of spam emails across the internet to distribute announcements of services and items to obtain commercial benefits. Spam detection became a challenge and most of the existing spam detection techniques are based on content-based filtering (CBF) which have many drawbacks. This paper aims to solve these CBF drawbacks by introducing designing, implementing, and evaluating a novel CBF approach for automatic detection of spam emails. The proposed approach is denoted as Intelligent and Hybrid Anti-Spam System. Its main theme is the neuro-fuzzy inference engine which has been applied successfully in a wide range of research. Initially, a normalization task is performed to remove any deceptive character from the text body. The preparation stage then examined the content of well-known datasets of spam emails to identify all the phrases that solely identify spam email. Our experiments show that IHASS could achieve very good accuracy level and works stably well. Keywords: Spam email detection, content-based filtering, neuro-fuzzy inference engine, kNN-based Evolving Neuro-Fuzzy Inference System. 1 INTRODUCTION E-mails are used in many aspects of our life including, education, banking, health and government services. Moreover, email is significantly potential for many businesses [1]. Spam is .
