Automated System for Detecting Python Code Plagiarism

This paper outlines a new approach to plagiarism detection between program codes performed in Python. The object of the study is a system for similarity detecting which combines different approaches of text and program analysis. The developed system was tested on programs created by third-year students of the educational programs “Applied Mathematics” and “Computer Security” as tasks in the disciplines “Computer Practicum” and “Python Programming Language”. In this report there is an overview of existing anti-plagiarism systems for program code. Moreover, methods for finding similarities in texts/ program codes were considered. As a result, the most effective methods were chosen to be implemented in the developed system. Created system excludes most of disadvantages which were found during critical analysis of analogues. Plagiarism checking system is written in Python and can be used by a teacher for work purposes to find clusters of students whose codes with a certain probability are similar. Two import methods were realized: files can be downloaded both from cloud and local storage.