Even the source code consists of data structures and algorithms, previous studies for plagiarism detection techniques on source codes do not consider the data structures. The data structure, however, has a resilience on plagiarism attacks, which impli...
Even the source code consists of data structures and algorithms, previous studies for plagiarism detection techniques on source codes do not consider the data structures. The data structure, however, has a resilience on plagiarism attacks, which implies that it tends not to change easily to the plagiarism attacks. In this paper, a plagiarism detection technique for source codes considering data structures is proposed. To show the usefulness of this technique, an experiment has been performed on 126 source codes submitted as homework results in an object-oriented programming course. When both the data structures and the algorithms of the source codes are considered, the precision and the F-measure score are improved 22.6% and 19.3%, respectively, than those of the case where only the algorithms are considered. This result shows that the method considering the data structures of the source code is effective to detect the plagiarized source code especially for improving the precision of the detection by reducing the number of false alarms.