Abstract: We describe a novel plagiarism detection system and its integration with an e-portfolio used in first year engineering teaching. The tool addresses an important issue arising from the decreasing barriers to information access. Academics know that information can support valuable learning experiences, but these experiences are diminished when students plagiarise by copying assignments and getting credit for work they have not done. While it is possible for academics to develop project-based activities to make it harder for students to plagiarise work from outside sources, some students will still copy work done by others within the same class, which can be especially difficult to detect within large cohorts. According to student feedback received while assessing an e-portfolio activity, we found that students were also concerned about plagiarism, and that they modify their approaches to learning based on this concern. We developed a plagiarism detection tool called Beagle, which uses an internal method (also known as collusion): whenever a student submits an assignment to the e-portfolio system, it is compared to those previously submitted by other students. Beagle measures the statistical similarity between students’ work using text mining methods. When a specific similarity threshold is reached, the work can be flagged as possible plagiarism or the system can automatically warn the student and request that they resubmit their work. In this paper we present the design of the system, a performance evaluation in terms of accuracy and execution time, and a description of its application integration capabilities through web services.
Citation: J. J. Garcia Adeva, Nicholas L. Carroll, and Rafael A. Calvo. “Applying Plagiarism Detection to Engineering Education”. Seventh International Conference on Information Technology Based Higher Education and Training (ITHET 2006), July 2006. IEEE Computer Society.
