Antis: Another Tool for Identical or Similar Code Detection.
Bogdan Ghiț, Matei Gruber, Octavian Voicu, Florin Pop · FCS · 2010
The goal of this project is to develop an application able to detect the plagiarism in programming assignments by identifying similar code sections. The evaluation process is realized by comparing source code in the abstract syntax trees (AST) representation. Detecting similarity between programs involves finding the maximum sub-graph that can be matched in both trees. The AST is a relational structure that provides a better understanding of the code. We may easily reduce this challenge to the well known tree matching problem, which has been proved to be NP-complete, as there is no deterministic algorithm able to compute the maximum match between two trees. In consequence, we developed several algorithms based on structural analysis of the abstract syntax trees and certain heuristics which are going to be explained in the following sections of the article. The aim of this work is to provide new approaches in the field of code source plagiarism detection that are going to increase both efficiency and accuracy of the matching process.