An XML Similarity Research For C Program

With the rapid development of the Internet, obtaining information becomes more convenient and quick, and plagiarism is easier than before. Due to the engineering practice, the teaching and assessment of computer-related majors almost entirely depend on the computers, which leads to the phenomenon that plagiarism and cloning become more and more common. Plagiarism become more and more serious, it affects not only the quality of teaching but also the improvement of student's quality. The similarity research of program code can found the program code existing suspected of plagiarism, and the accuracy and objectivity of detection can be guaranteed.On the basis of analysis of the domestic and foreign similarity research of program code, this paper proposes a similarity research methods for C programs based on XML. Firstly, the key structure that can represent the procedural structure from C language is picked out, summarizing the means of plagiarism of possible existence of key structure, unifying the programming style of source program, and formating the source program. Secondly, determining the specific extraction information, according to the different key structure and corresponding means of plagiarism, the different algorithm of generating an XML text from program code is designed, it is said that using XML text represent C program. Thirdly, according to the different key structure and corresponding means of plagiarism, the different similarity comparison algorithm of XML text line is designed to calculate the similarity of XML text line. Finally, according to the similarity of XML text line, the procedural similarity comparison algorithm is designed to calculate the similarity of two programs, and then judging the plagiarism of pairs of program。On the basis of the above-mentioned theory research, the author develop a experimental system of plagiarism detection for C program based on XML, and then choose different validation methods to verify the efficiency and reliability of the experiment system. The experiment prove that the similarity research methods for C programs based on XML can effectively detect the plagiarism of program existing kind's means of plagiarism, and the system can effectively detect the following common plagiarism:function, structures, pointers, arrays, replacing control structure with equivalent's control structure, and the equivalent substitution of expression.