Data mining systems used to help students stay safe, find experts
July 4, 2010
Filed under News
Student groups from the Data Sciences Summer Institute (DSSI) presented different digital systems — created with hopes of making relevant information more accessible to people not experienced in information technology — on Friday.
“Most of the data today is unstructured,” Dan Roth, professor in Computer Science and director of DSSI, said before the student presentations. “And the question is, ‘How can we deal with this huge amount of information as if we were organizing some database with its own schema and queries?’”
The three groups — consisting of 22 students from the University as well as other schools throughout the country — had four weeks to research and prepare for Friday’s presentations, which took place for an audience of full seats.
Partially funded by the U.S. Department of Homeland Security, the DSSI’s mission “is to develop diverse human resources to enhance the scientific research, education and government workforce in Data Science disciplines,” according to its website. A total of 170 students applied to the institute, though only 22 were admitted.
Using different types of data mining (where mathematical algorithms are used to probe through websites and identify information relevant to key words or other specifications) the groups came up with systems that search for three different things: experts in their field, similar images and locations of recent criminal activity.
Group members Matt Gornick and Blaine Fahey, who worked on the crime location project, came out with Illinicrime.com earlier this year. The website, now available as an application for iPhones, tracks the location of crimes committed on campus using data from University Police Department’s Crime Alert notices.
Their presentation Friday included a demo of a system that tracks local crime in a similar way, but a larger number of crime reports — this time from the Champaign Police Department — were used instead. With the system, users are able to type a certain crime into the search bar, and the locations of the recently committed crimes are displayed on a Google Maps image, along with other aggregate information about the specific crimes.
“Really now, you can take the crime information when you’re going out and if you’re coming back late at night knowing where crime has happened in the past,” said Gornick, a recent graduate in Computer Science. “And hopefully, we would be able to provide some information how to get back to your apartment safe.”
Group member Evan Misshula, a doctoral student at the City University of New York’s John Jay College of Criminal Justice, said it is important that these types of applications are available to everyday people at convenient times.
“The reason, obviously, that we want to render on mobile devices is you make your decision about where to go and what to do not always when you’re at your desk, but when you’re out in the world,” Misshula said.