Okesola, J. O. and Ojo, F. O. and Longe, O. B. (2015) On the use of Locality for Improving SVM-Based Spam Filtering. Computing, Information Systems, Development Informatics & Allied Research Journal, 6 (2). pp. 7-12.
![[img]](http://eprints.covenantuniversity.edu.ng/style/images/fileicons/application_pdf.png) |
PDF
Download (148kB) |
Abstract
Recent growths in the use of email for communication and the corresponding growths in the volume of email received have made automatic processing of emails desirable. In tandem is the prevailing problem of Advance Fee fraud E-mails that pervades inboxes globally. These genres of e-mails solicit for financial transactions and funds transfers from unsuspecting users. Most modern mail-reading software packages provide some forms of programmable automatic filtering, typically in the form of sets of rules that file or otherwise dispose mails based on keywords detected in the headers or message body. Unfortunately programming these filters is an arcane and sometimes inefficient process. An adaptive mail system which can learn its users’ mail sorting preferences would therefore be more desirable. Premised on the work of Blanzieri & Bryl (2007), we proposes a framework dedicated to the phenomenon of locality in email data analysis of advance fee fraud e-mails which engages Support Vector Machines (SVM) classifier for building local decision rules into the classification process of the spam filter design for this genre of e-mails.
| Item Type: | Article |
|---|---|
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science |
| Depositing User: | Mr Adewole Adewumi |
| Date Deposited: | 19 Mar 2018 13:24 |
| Last Modified: | 19 Mar 2018 13:24 |
| URI: | http://eprints.covenantuniversity.edu.ng/id/eprint/10410 |
Actions (login required)
 |
View Item |
