Beyond Spam: Spit and gray mail

Also at Georgia Tech, researchers are looking into how to prevent Spit, or Spam over Internet Telephony. "CallRank: Combating Spit using Call Duration, Social Networks, and Global Reputation," discusses the use of a system called CallRank that uses call duration to establish social-network linkages and reputations for callers. Based on this information, VoIP users can decide whether a caller is legitimate or not, they say.

Other research projects include rating detectors of gray mail, or messages that some people may consider spam, while others would not. In a paper called "Improve Spam Filtering by Detecting Gray Mail," Microsoft researchers propose three ways to detect gray mail and then compare their performance.

Out of IBM Research comes a paper that examines the efficacy of combining global and personal antispam filtering systems. The paper, called "Combining Global and Personal Anti-Spam Filtering," examines the advantages of using personally trained antispam filters that do better at statistical text learning, because they are tuned to the individual user's unique e-mail aspects, the researcher says.

Yet classifiers learned for a large number of users can leverage the information provided by each user across the whole group. The paper discusses how combining the two approaches improves overall spam detection.

Researchers aren't just looking into the technology behind spam, but also the economics. A project headed by researchers from CMU and the University of California at San Diego called "An Inquiry into the Nature and Causes of the Wealth of Internet Miscreants" studies the underground economy that the Internet has fostered. Examining activities such as credit-card fraud, identity theft, spamming, phishing, online credential theft, and the sale of compromised hosts, the paper attempts to explain how Internet abuse once considered a hobby has now turned into a multimillion-dollar business.