Research And Realization Of String Matching Algorithm For Detecting Plagiarism In Program Code

Plagiarism in student program tasks makes bad effect to teaching. Detecting plagiarism tasks and punishing the students who plagiarize others is helpful to fall the number of plagiarism, so this can improve the study effect in programming course and inspire the students working hard. However, detecting plagiarism manually and finding suspicious program task is troublesome and invalid. Application the detecting software can realize the automatic disposal, and help teacher to a large degree. In other ways, the approach also can apply for detecting software edition and copyright.This dissertation introduces the theory about detecting for program code, including the realization, key technology, detecting effect and some applied system. This dissertation also argues the usability of application the detection system on detecting student tasks.String matching arithmetic is an important approach in tokens matching of detecting plagiarism for program code, and the characteristic of program code demand the matching result should not change because of the pattern length or the token position or other factors. The dissertation mainly studies the string matching arithmetic that is appropriate to detecting plagiarism for program code. Analyzing several arithmetic that has taken by some detecting system, such as the dynamic programming approach, Hechel arithmetic and GST arithmetic, and summarize these arithmetic's merit and shortage. After the compare of above arithmetic, the GST arithmetic is more appropriate. So this dissertation research and realize the GST arithmetic and the improvement arithmetic of GST—RKR-GST arithmetic. The dissertation takes correlative experiment for GST and RKR-GST arithmetic. The effects show that these tow arithmetic are appropriate to program tokens match.