Campus News

Computer science plagiarism program developed by SJU graduate

June 25, 2010

By Mike Killeen

Jeremy Iverson and Deng Hanjie share dinner together while working on the research project.

When classmates heard that Jeremy Iverson was working on a computer program to detect plagiarism in class, he heard about it.

"Everybody gives me a little bit of a hard time at first," said Iverson, from Champlin, Minn., who graduated May 9 with a degree in computer science from Saint John's University, Collegeville.

But the professors in the computer science department at the College of Saint Benedict and Saint John's University might feel differently. Come this fall, the detection system that Iverson helped create will be used by professors to detect plagiarism in the introductory computer science class for non-majors.

Known as "VOCS" (Versatile Online Clustering System), the program will search for plagiarism in projects submitted to the professors.

"The initial implementation has been done for the Visual Basic programming language which we use for our non-majors course," said James Schnepf, associate professor of computer science and chair of the department, who was also Iverson's advisor. "Visual Basic was chosen as the first language because we have a large number of students who take the non-major course. Students in the course implement a project of their choosing to demonstrate their mastery of the material. This presents a problem as students could potentially copy a project from a previous semester."

"What we're trying to do is detect these lifted ideas," Iverson said.

Project began with another SJU student

Two years ago, SJU student Joe Degiovanni laid the groundwork for Iverson's project. Working with Imad Rahal, assistant professor of computer science at CSB and SJU, Degiovanni investigated various options for finding plagiarism and developed the initial algorithm for the approach used by Iverson, Schnepf said.

In the summer of 2009, Iverson participated in the China summer research program. He and his Chinese student partner, Deng Hanjie, extended Degiovanni's work and implemented his algorithm.

"Jeremy did studies of plagiarized programs to determine the best approach for determining if two sets of code were similar enough to be considered plagiarism," Schnepf said. "He continued his work as his thesis during the 2009-10 academic year."

How it works

Iverson created a database of about 650 previously submitted programming projects. When a new project is submitted, professors will submit the program to VOCS. Various steps are conducted to see if there are similarities to other programs.

"If there are too many similarities, the program is flagged for plagiarism," Schnepf said.

"This is going to help the professor weed out projects that we know for sure are not plagiarism - you don't have to worry about those," Iverson said. "VOCS is going to say, take a look at these."

Iverson, who is headed off to graduate school at the University of Minnesota, says the system will help students learn and complete the well-roundedness of a liberal arts education.

"If you plagiarize your way through computer science class, you're losing that well-roundedness," Iverson said. "I think the teachers view it as a benefit not only to themselves to reduce the amount of work they do, but I think they honestly care about the students, and they want to see them learn. They enjoy computer science, and they want to help these students and not just get through it and get a grade."