Hoodwinking spam email filters

Authors:
Wanli Ma;Dat Tran;Dharmendra Sharma;Sen Li
Affiliations:
School of Information Sciences and Engineering, University of Canberra, Australia;School of Information Sciences and Engineering, University of Canberra, Australia;School of Information Sciences and Engineering, University of Canberra, Australia;School of Information Sciences and Engineering, University of Canberra, Australia
Venue:
CEA'07 Proceedings of the 2007 annual Conference on International Conference on Computer Engineering and Applications
Year:
2007

Citing 4
Cited 0

A Memory-Based Approach to Anti-Spam Filtering for Mailing Lists

Information Retrieval
A LVQ-based neural network anti-spam email approach

ACM SIGOPS Operating Systems Review
Adaptive Spam Filtering Using Dynamic Feature Space

ICTAI '05 Proceedings of the 17th IEEE International Conference on Tools with Artificial Intelligence
Support vector machines for spam categorization

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

Many spam email filters have been proposed, however spammers regularly find new ways of hoodwinking those filters. Most of those filters are text based and hence spammers try to conceal the text which reveals the spam nature of an email. In order to investigate the ways spammers are using, we consider a large set of spam emails and found that we can classify these emails into 5 categories which are text based, obfuscating, image based, HTML tags, and non-English. We counted the percentage of spam emails in each category and then used a sample spam filter to evaluate the effectiveness of the filter on each of the categories. The TREC Spam Filter Evaluation Toolkit was used in our evaluation.