A TWO-STEP IN-CLASS SOURCE CODE PLAGIARISM DETECTION METHOD UTILIZING IMPROVED CM ALGORITHM AND SIM

摘要

Source code plagiarism is becoming one of the most serious problems in academia. There have been many proposed methods that attempt to detect source code plagiarism in programming classes. Most of them extract algorithmic features from the source code and measure the similarity between them. These methods show high levels of accuracy in evaluation experiments, and however, it is concerning that the similarity detected by the methods might not be caused by plagiarism. As a result, we propose a method called the CM Algorithm, which utilizes a student's coding style, the way the student writes source code, to check whether the source code submitted by the student was produced by him/her. In this paper, we propose a combined method that measures the similarity between source codes by using SIM [7], one of the well-known in-class source code plagiarism detection systems, and then checks the outputs of SIM against our improved CM Algorithm. The new method is expected to reduce false positives in plagiarism detection systems. This paper also gives a detailed explanation of the improved CM Algorithm, which assumes fluctuations in the source code produced by a student's coding style.