Abstract

Spammers collect email addresses from internet using automated programs known as bots and send bulk SPAMS to them. Making the email address difficult to recognize for the bots (obfuscate) but easily understandable for human users is one of the effective way to prevent spams. In this paper, we focus on evaluating the effectiveness of different techniques to obfuscate an email address and analyze the frequency at which spam mails arrive for each obfuscation technique. For this we employed multiple web crawlers to harvest both obfuscated and non-obfuscated email addresses. We find that majority of the email addresses are non-obfuscated and only handful are obfuscated. This renders majority of email users fall prey to SPAMS. Based on our findings, we propose a natural language processing (NLP)-based obfuscation technique which we believe to be stronger than the currently used obfuscation techniques. To analyze the frequency of arrival of spam mails in an obfuscated mail, we posted obfuscated email addresses on popular websites (social networking and ecommerce sites) to analyze the number of spams received for each obfuscation technique. We observe that even simple obfuscation techniques prevent spams and obfuscated mails receive less spam mails than the non-obfuscated ones.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call