BYU undergraduate computer science student Tobias Kin Hou Lei has created a bird’s-eye view of crime by analyzing tweets across the nation. But he wasn’t just winging it — Lei participated in a six-week long research project aiming to take crime prevention into the future.
While at the Data Sciences Summer Institute (DSSI), a research program at the University of Illinois Urbana-Champaign, Lei and students from multiple universities developed a new program using Twitter. Their program takes updates from individual users (i.e. tweets) and uses machine-learning and data-mining approaches to classify them in relation to different crime groups. Once they’ve been sorted, they are entered into a database for analysis. Since July, 1.2 billion tweets have been classified and entered into the database, with more being added everyday.
The data can be used to compare changes in crime that show differences over time between specific cities’ crime rates and even to generate a live map showing when and where people are tweeting about crime. Yahoo! honored Lei’s group as having the best DSSI project this year, and each student in the group received a scholarship award.
A program like the one designed by Lei’s research group has a variety of potential uses. Lei was most excited about using the program as an early warning system to help prevent coordinated crime efforts or terrorist attacks.
“If we get a lot of data from a resource like Twitter and can find a pattern from the data, then we can do something about it,” Lei said.
Even without the scholarship money or an opportunity to capitalize on the project, Lei said the research institute was a great experience. He was able to get to know another university, gain useful research experience and work alongside masters’ and PhD students. After working at the institute, Lei said he knew more about what options would be available to him after he graduates. He also said that he had a better idea of what he wants to focus on as a graduate student.
Undergraduate students like Lei who participate in research will be more prepared for the field when they graduate, and Lei said he thinks all students should try to learn from real research projects and from other universities.
“I learned how to work on a real research project,” Lei said. “Usually you just work on a small project to get a grade, and after class you just throw it away. But actually, [students] can do something like this and sell the project. It’s very different than a school project.”
Lei and two other students on the project have submitted a paper and been accepted to present it at the International Conference on Data Engineering. Lei is currently working under Dr. Eric Ringger as a research assistant, and will graduate in April of 2012.