Methods and Realization of Intelligent Visual Detection on Plagiarized C Source Code
|School||Dalian University of Technology|
|Course||Control Theory and Control Engineering|
|Keywords||Plagiarism code Similarity Visualization Attribute vector model SMR method|
The source code plagiarism detection , the similarity is an important application of the identification code . Objectivity and authenticity of the program design for a computer programming language teaching assessment , looking intelligently identify plagiarism code of the C program , to achieve high efficiency , easy operation , visual detection target . Plagiarism code type and identification methods plagiarism code identification and detection methods of student work from two aspects . First the code of student work program features a simple, short code length data structure and a single type , but the large number of typical characteristics of the the characteristic attribute vector space model - based clustering method . The method by extracting the the property attributes vector space model able to represent the characteristics of the source code , the introduction of the vector space clustering algorithm divided the plagiarism class , in a relatively short period of time quickly identify the source code of the suspected plagiarism . Secondly , on the basis of access to suspected plagiarism source file , in order to obtain a more intuitive visual detection effect , from the identification of the two source files copying a regional perspective , this paper, the recognition method based on the structure metric SMR match through two over structural analysis of the source file into function code segment RKR-GST string matching algorithms , and the introduction of the two source files , the level of the word segment match , thereby identifying a similarity higher than a threshold value of the code segment positioning plagiarism region visualization of the detection results. Based on the above identification method , design visualization intelligent detection system for copying and C source code , build plagiarism code detection evaluation system , to provide reliable and objective evaluation based on test results . The tests show that the detection system capable of detection and identification of common types of plagiarism code for the operating procedures more common high similarity the plagiarism program effects significantly . The system is to be applied also actual course examination can accurately identify more than 80% of the plagiarism program , played a good role in improving the quality of teaching and improve student learning outcomes , demonstrated a strong practical value .