At the Georgia Institute for Technology, new software has been developed that can identify spam before it hits a mail server. SNARE (Spatio-temporal Network-level Automatic Reputation Engine), rates every incoming e-mail based on a single packet of data cross referenced with new criteria researchers put together. The researchers behind the project believe that the automated system puts a lesser strain on the network and minimizes the need for human intervention while achieving the same accuracy as any traditional spam filter.
25 million e-mails were analyzed after their collection by TrustedSource, an online service developed by McAfee to collate data on trends in malware and spam. With this data, the Georgia Tech researchers were able to discover several characteristics that could be used to efficiently identify junk mail.
The research team plans to present their work on SNARE at the Usenix Security Conference in Montreal next month. Courtesy of computerworld.com