November 20, 2012
New Spam Filter Tools Could Keep Your Inbox Clean
Michael Harper for redOrbit.com — Your Universe Online
Spam. It´s loathed by all who aren´t responsible for sending it out, it´s crowding inboxes worldwide, and it refuses to stand still.Nearly every modern mail client, online or off, features some sort of spam filter which automatically siphons off these unsolicited messages into another folder which we hopefully never have to see.
Spam doesn´t appear from thin air, however, and some person somewhere is making money on these messages. As such, they do all they can to make sure those messages not only land in your inbox, but that you are lured into clicking on them, and maybe, just maybe, handing over some cash or, at the least, your personal information.
Researchers at Concordia University´s Institute of Information Systems Engineering are working towards eliminating spam altogether by moving past the early days of text-based messages and into the new techniques of using images to compose a message. Modern day spam filters look at the text of the message to determine if a message should be blocked, but these days spam has gotten clever, using images to get around these blockers.
Doctoral candidate Ola Amayri and Nizar Bouguila, thesis supervisor for this research, have come up with a solution which they say is able to catch and filter out even the trickiest of messages.
“The majority of previous research has focused on the textual content of spam emails, ignoring visual content found in multimedia content, such as images. By considering patterns from text and images simultaneously, we´ve been able to propose a new method for filtering out spam,” says Amayri in a statement. Amayri´s research is set to be published online and will even be discussed at a series of international conferences.
Modern day spam uses images to obscure the text which would normally set off a spam filter. It often uses symbols instead of words and even layers backgrounds on top of one another to cover up the spammy message.
“Our new method for spam filtering is able to adapt to the dynamic nature of spam emails and accurately handle spammers´ tricks by carefully identifying informative patterns, which are automatically extracted from both text and images content of spam emails,” says Amayri.
The Montreal team were able to create what they hope will be an effective spam tool by conducting experiments on traditional spam filters. Then, the team used some new techniques and taught these tools how to recognize patterns in spam, moving past text-based patterns but image-based patterns as well.
Amayri and team tested this new filter on English-language spam messages, but say they´ll easily be able to change the filter to work in other languages as well.
This new spam tool is still in its development phase. However, Amayri and Bouguila are working on a plug-in for SpamAssassin, a popular open-sourced spam filtering tool. By creating this plug-in, Amayri hopes others will take continue their research and work together to bring an end to spam for good.
“Spammers keep adapting their methods so that they can trick the spam filters,” says Amayri. “Researchers in this field need to band together to keep adapting our methods too, so that we can keep spam out and focus on those messages that are really important.”