Research And Application Of Plagiarism Detection Technology Based On Code Style Classification

With the development of network information and the convenience of resource acquisition,people's information acquisition habits have changed.People are more used to obtaining electronic resources through network search,and at the same time,the modification of electronic resources is more convenient.For most electronic resources without any protection measures or author information(such as electronic signature and password),the following problem is the ownership of electronic resources or plagiarism discrimination.In computer programming education,source code plagiarism occurs from time to time.At the same time,online programming platform and programming competition platform also have many such phenomena.Students' learning from others' code or even copying others' code not only affects the cultivation of students' ability,but also violates academic integrity.Code plagiarism in programming competition platform seriously affects the fairness of the competition and the credibility of the platform.At present,although there are some code plagiarism detection systems,the recognition of type-3 and type-4 clones is not accurate enough,and in the case of large amount of data,the efficiency of the system needs to be improved.Therefore,it is necessary to develop an efficient code plagiarism detection system which can effectively identify type-3 and type-4.Based on the in-depth study of code similarity detection,this paper proposes a multi feature joint code similarity detection model Jat bilstm for the recognition of type-3 and type-4 code clones.Aiming at the problem of system efficiency,this paper proposes a code plagiarism detection scheme combining user programming style classification and Jat bilstm algorithm,which identifies the anonymous author of the code through code style matching,so as to reduce the code comparison set and improve the detection efficiency.Based on the above research results,a code plagiarism detection system based on code style classification is designed and implemented,which is applied to OJ platform.The experimental results show that the accuracy and F1 score of the Jat bilstm similarity detection model designed in this paper are better than those of the control experiment in the codeojpy dataset,and the comprehensive performance is the best;In codeoj dataset,the accuracy,recall and F1 score are better than the control experiment,and the recognition accuracy of type-3 and type-4 is better than the control algorithm.The proposed code plagiarism detection system based on code style matching can accurately identify similar code pairs and effectively improve the efficiency of the system.