ES‐Plag: Efficient and sensitive source code plagiarism detection tool for academic environment

Abstract

Source code plagiarism detection using Running‐Karp‐Rabin Greedy‐String‐Tiling (RKRGST) is a common practice in academic environment. However, such approach is time‐inefficient (due to RKRGST's cubic time complexity) and insensitive (toward token subsequence rearrangement). This paper proposes ES‐Plag, a plagiarism detection tool featured with cosine‐based filtering and penalty mechanism to handle aforementioned issues. Cosine‐based filtering mitigates time‐inefficiency by excluding non‐potential pairs from RKRGST comparison; while penalty mechanism mitigates insensitivity by reducing the number of matched tokens with the number of matched subsequences prior similarity normalization. In addition to issue‐solving features, ES‐Plag is also featured with project‐based input, colorized adjacency similarity matrix, matched token highlighting, and various similarity algorithms (e.g., Cosine Similarity and Local Alignment). Three findings can be deducted from our evaluation. First, cosine‐based filtering boosts up time efficiency with a trade‐off in effectiveness. Second, penalty mechanism enhances sensitivity even though its improvement in terms of effectiveness is quite limited. Third, ES‐Plag's features are beneficial for examiners.

TF-IDF Inspired Detection for Cross-Language Source Code Plagiarism and Collusion

“…The first research question was addressed by comparing the proposed technique (with internal IDF indexes on board, referred as idfw ) to common techniques in academia [49] based on the similarity degree of the suspected pairs and MAP. These common techniques in academia were slightly modified to deal with cross-language source code plagiarism and collusion.…”

“…Source code plagiarism and collusion occur when a person steals another person's source code and claims it as theirs [11,19,47]. Several automated detection techniques have been proposed to capture such a misbehavior [49]. They commonly compare the source code files in a pairwise manner in which pairs with high similarity degrees are considered to be suspicious.…”

“…Structure-based techniques are argued to be at least as effective as attribute counting-based techniques [52], even though they are considerably slower [49]. These techniques rely on various structures for comparison.…”

“…Since attribute counting-based and structure-based techniques have their own characteristics and combining them may lead to more benefits [25], hybrid techniques have been introduced. A structure-based technique can be more time-efficient if the source code pairs are limited to those that share a high attribute counting-based similarity degree [9,49]. On the contrary, an attribute counting-based technique can be more effective if its results are combined with a structure-based technique's result [15,40,45].…”

A Language-Independent Library for Observing Source Code Plagiarism

“…This could lead to several problems, especially those which are related to authorship and creative works. Hence, several automated plagiarism detection tools dedicated to code domain (e.g., JPlag [2], ES-Plag [3], and IC-Plag [4]) have been proposed. Using those tools, such an illegal behaviour can be detected with limited human effort.…”