In programming courses there are various ways in which students attempt to cheat.
The most commonly used method is copying source code from other students and making
minimal changes in it, like renaming variable names. Several tools like Sherlock,
JPlag and Moss have been devised to detect source code plagiarism. However, for
larger student assignments and projects that involve a lot of source code files
these tools are not so effective. Also, issues may occur when source code is given
to students in class so they can copy it. In such cases these tools do not provide
satisfying results and reports. In this study, we present an improved process
model for plagiarism detection when multiple student files exist and allowed
source code is present. In the research in this paper we use the Sherlock
detection tool, although the presented process model can be combined with any
plagiarism detection engine. The proposed model is tested on assignments in
three courses in two subsequent academic years.
plagiarism detection, source-code, process model, programming assignments.
To preview full
article text in PDF format click here
could obtain free Acrobat Reader from Adobe