Title: E-mail spam classification using S-cuckoo search and support vector machine
Authors: T. Kumaresan; C. Palanisamy
Addresses: Department of Computer Science and Engineering, Bannari Amman Institute of Technology, Sathyamanglam, Tamilnadu, India ' Department of Information Technology, Bannari Amman Institute of Technology, Sathyamangalam, Tamilnadu, India
Abstract: Today's internet world email spam becoming a major problem to the internet users. The main intention of this paper is to design a technique for email spam classification using modified cuckoo search called as stepsize-cuckoo search (SCS) and support vector machine. Optimised feature set is identified with the use of SCS algorithm. Once the best feature set is identified through SCS, the spam classification is done using the support vector machine. For the effectiveness of classification, we have used three different kernels such as linear, polynomial and quadratic. To evaluate the proposed email spam classification, we have used evaluation metrics such as sensitivity, specificity and accuracy. From the results, it clearly shows that our method has shown higher accuracy than the original cuckoo with SVM for spam base datasets.
Keywords: step size; S-cuckoo search; SVM; e-mail; legitimate; spam; training; testing; feature set.
DOI: 10.1504/IJBIC.2017.083677
International Journal of Bio-Inspired Computation, 2017 Vol.9 No.3, pp.142 - 156
Received: 18 May 2015
Accepted: 11 Jul 2015
Published online: 19 Apr 2017 *