A method for detecting suspected code plagiarism based on a random forest model

Abstract

The invention relates to a method for detecting suspected code plagiarism based on a random forest model. The method can be divided into two stages from a large aspect, the first stage is a feature extraction stage, and the second stage is a model training and prediction stage. The method collects the relevant characteristic data of the two codes to be detected, the characteristic data of the users to be detected and the attributes of the related topics, and modeling is carried out by introducing a random forest algorithm to obtain whether the current users are suspected of plagiarism or not.

Description

technical field [0001] The invention relates to a method for detecting suspected code plagiarism based on a random forest model, which belongs to the application field of machine learning technology. Background technique [0002] Nowadays, with the rapid development of computer technology, a large number of people have poured into the computer industry, and a series of honesty problems brought about by the surge of employees cannot be ignored. As small as programming papers in computer science, as large as key software engineering products, the phenomenon of cloning and plagiarism in program codes is becoming more and more serious. The plagiarism methods mainly include the following: (1) unchanged; (2) modifying comments; (3) modifying identifiers; (4) adjusting variable positions; (5) process combinations; (6) adjusting statement positions; ( 7) Adjust the control structure logic, etc. In addition, some scholars have conjectured and constructed other means of plagiarism i...

AI Technical Summary

Problems solved by technology

However, as a traditional neural network, the BP neural network has only one layer in its hidden layer. In the process of training the feature set, the weight value of each feature output by the neural network tuning cannot achieve the ideal prediction effect. also to be verified

Moreover, among the 7 eigenvalues ​​collected by the BP neural network model of the invention, some data are unstable, such as code style similarity, statistical attribute similarity and other characteristics. These characteristics themselves have relatively small differences in terms of program syntax. In addition, the amount of code required for algorithm practice is small, so in the process of model classification, the final realization of the distinction is not large, and the performance of the model cannot be better played

Images