VANCOUVER — Deep within the recesses of the Internet, extremists discuss and plot terrorist acts. But new mathematical tools that combine web crawling techniques, sophisticated algorithms and human expertise are gaining access to this “dark side” of the Web and may help predict and prevent violence.
Researchers engaged in the “Dark Web Project,” a program started partly in response to the 9/11 terrorist attacks, have developed methods for tracking the spread of dangerous ideas through certain rogue and jihadi Web forums. Using a mathematical model known as SIR, used by epidemiologists to describe the transmission of disease, researchers have determined that the infection rate for becoming a suicide bomber is 2 in 10,000, Hsinchun Chen of the University of Arizona in Tucson reported February 18 at the annual meeting of the American Association for the Advancement of Science.
“Violence in social media is infections of the mind,” said Chen.
The Dark Web Project, housed at the University of Arizona, collects information from blogs, forums and other websites from hidden realms of the Web. Search engines typically explore only what’s known as the publicly indexable web. The invisible Web, which includes these Dark Web forums, is estimated to contain 500 times as much information as the surface web.