Software plagiarism detection method based on reduced shortest path birthmarks

Abstract

The invention discloses a software plagiarism detection method based on reduced shortest path birthmarks, which comprises the following steps: 1) statically disassembling a to-be-analyzed program to generate a to-be-analyzed function table; 2) for each function, searching a shortest path set from an entrance node of the control flow graph to other nodes, performing path reduction and abstraction,and constructing a reduced shortest path birthmark; 3) calculating the similarity between the reduced path birthmarks by using fuzzy hashing to serve as a measure of the similarity between functions;4) grouping the functions, and generating a function mapping table through maximum weighted bipartite graph matching; 5) calculating the similarity between the original program and the announced program, carrying out plagiarism judgment, and outputting a detection result. Through path searching and abstraction, the code confusion resisting capacity of the method is improved, and through fuzzy hashing and grouping matching, it is guaranteed that the method is applied to scalability of a large-scale real binary program.

Description

technical field [0001] The invention belongs to the technical field of program feature mining and software plagiarism detection, in particular to a software plagiarism detection method based on reducing the shortest path birthmark. Background technique [0002] The vigorous development of open source software communities such as GitHub, SourceForge, and social programming sites such as CodeShare has brought prosperity to the software industry; the accompanying problem of software plagiarism has become increasingly serious, and abuse of other people's code is not uncommon. On the one hand, there is no lack of premeditated plagiarism driven by economic interests. In 2015, APICloud, a product of Grapefruit Technology, was suspected of plagiarizing a large number of codes of Digital Paradise DCloud and fell into an infringement case. The judicial appraisal report confirmed that it not only directly misappropriated DCloud’s dll files, but also copied DCloud’s functions such as ru...

AI Technical Summary

Problems solved by technology

On the one hand, there is no lack of premeditated plagiarism driven by economic interests

In 2015, APICloud, a product of Grapefruit Technology, was suspected of plagiarizing a large number of codes of Digital Paradise DCloud and fell into an infringement case. The judicial appraisal report confirmed that it not only directly misappropriated DCloud’s dll files, but also copied DCloud’s functions such as running on the real machine and viewing while modifying. A large number of source codes; the recent "Red Core Disturbance" is even more noisy, claiming that the Red Core browser, which independently developed the domestic kernel, was exploded just to change the shell of Google Chrome; on the other hand, it was unintentional plagiarism

Most of the existing methods are difficult to resist the interference caused by different compilers and optimization options due to insufficient consideration of program semantics and behavior when constructing birthmarks

[0006] 3. The scalability of existing plagiarism detection methods is not good, and it is often difficult to deal with large-scale programs in reality

Most of the existing birthmark technologies mainly solve the problem of overall plagiarism. Due to the coupling and correlation between codes, it is often difficult for codes suspected of plagiarism to be packaged and run independently, which greatly limits the use of dynamic birthmark technologies with stronger detection capabilities. Applications in Local Plagiarism Detection

Images