Some very bad news for all the internet trolls, spoofers, phishers, spammers, cyber-bullies, cyber-stalkers, email bombers, and child pornographers. The days of your anonymous email address shielding your true identity are numbered.
Lesley Ciarula Taylor of The Toronto Star writes about a new data-mining method, developed by engineers and computer scientists at Concordia University, which has cracked the code for tracing anonymous emails:
For the first time, said data-mining expert Benjamin Fung, analysts have used the complex algorithms and almost imperceptible human quirks that make up the concept of “frequent pattern” to work out each person’s unique email fingerprint or “write-print.”
“The people who wrote the email don’t even recognize what they are doing,” Fung told the Star. “One of the features we break down is vocabulary richness. That would be hard to increase quickly.”
Other telltale evidence of the mystery writer can come from common grammatical mistakes, an unconscious extra space between each paragraph or patterns in punctuation.”
Fung acknowledges that the “write-print” method is not fool-proof yet, noting that:
It does have its limits. An investigator can capture the Internet provider (IP) of an email and thus trace it to one house or one office. From there, the “frequent pattern” system can weed out other users and find the culprit.
This new “write-print” data-mining method will likely be instantly popular among law enforcement types limited by traditional forensic hand-writing and document analysis, particularly in the areas of cyber-crime and cyber-terrorism. For civil trial lawyers, the benefits are immediately apparent. Beforehand, even if you were able to trace an email address from an ISP using an IP address at a specific location (e.g. a house or office), you still may not have been able to eliminate the possibility that someone else with computer access was the culprit. Now you may be able to winnow down the email author’s identity based upon his or her unique writing style, syntax, semantics and typing quirks.