An Approach to Source-Code Plagiarism Detection and Investigation Using Latent Semantic Analysis

Cosma, Georgina; Joy, Mike

doi:10.1109/tc.2011.223

Cited by 144 publications

(93 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…A number of tools have been developed specifically for computer science courses by addressing the unique challenges of examining code, see e.g. Lancaster and Culwin (2004), Cosma and Joy (2012), or Durić and Gašević (2013). Occasionally students raise concerns regarding the privacy and security of their intellectual property with these kinds of tools, or with the efficacy of the tool itself (Jones & Moore, 2010).…”

Section: Detection and Disciplinementioning

confidence: 99%

An Analysis of the Effect of Mobile Learning on Lebanese Higher Education

Jabbour¹

2014

Informatics in Education

View full text Add to dashboard Cite

We present an overview of the nature of academic dishonesty with respect to computer science coursework. We discuss the efficacy of various policies for collaboration with regard to student education, and we consider a number of strategies for mitigating dishonest behaviour on computer science coursework by addressing some common causes. Computer science coursework is somewhat unique, in that there often exist ideal solutions for problems, and work may be shared and copied with very little effort. We discuss the idiosyncratic nature of how collaboration, collusion and plagiarism are defined and perceived by students, instructors and administration. After considering some of the common reasons for dishonest behaviour among students, we look at some methods that have been suggested for mitigating them. Finally, we propose several ideas for improving computer science courses in this context. We suggest emphasizing the intended learning outcomes of each assignment, providing tutorial sessions to facilitate acceptable collaboration, delivering quizzes related to assignment content after each assignment is submitted, and clarifying the boundary between collaboration and collusion in the context of each course. While this discussion is directed at the computer science community, much may apply to other disciplines as well, particularly those with a similar nature such as engineering, other sciences, or mathematics.

show abstract

Section: Detection and Disciplinementioning

confidence: 99%

An Analysis of the Effect of Mobile Learning on Lebanese Higher Education

Jabbour¹

2014

Informatics in Education

View full text Add to dashboard Cite

show abstract

“…These techniques show positive results and open further possibilities in this research area. Examples of these techniques include Software Bertillonage (Davies et al 2013), Kolmogorov complexity (Li and Vitâanyi 2008), Latent Semantic Indexing (LSI) (McMillan et al 2012), and Latent Semantic Analysis (LSA) (Cosma and Joy 2012).…”

Section: Code Similarity Measurementmentioning

confidence: 99%

A comparison of code similarity analysers

2017

View full text Add to dashboard Cite

Copying and pasting of source code is a common activity in software engineering. Often, the code is not copied as it is and it may be modified for various purposes; e.g. refactoring, bug fixing, or even software plagiarism. These code modifications could affect the performance of code similarity analysers including code clone and plagiarism detectors to some certain degree. We are interested in two types of code modification in this study: pervasive modifications, i.e. transformations that may have a global effect, and local modifications, i.e. code changes that are contained in a single method or code block. We evaluate 30 code similarity detection techniques and tools using five experimental scenarios for Java source code. These are (1) pervasively modified code, created with tools for source code and bytecode obfuscation, and boiler-plate code, (2) source code normalisation through compilation and decompilation using different decompilers, (3) reuse of optimal configurations over different data sets, (4) tool evaluation using ranked-based measures, and (5) local + global code modifications. Our experimental results show that in the presence of pervasive modifications, some of the general textual similarity measures can offer similar performance to specialised code similarity tools, whilst in the presence of boiler-plate code, highly specialised source code similarity detection techniques and tools outperform textual similarity measures. Our study strongly validates the use of compilation/decompilation as a normalisation technique. Its use reduced false classifications to zero for three of the tools. Moreover, we demonstrate that optimal configurations are very sensitive to a specific data set. After directly applying optimal configurations derived from one data set to another, the tools perform poorly on the new data set. The code similarity analysers are thoroughly evaluated not only based on several well-known pair-based and query-based error measures but also on each specific type of pervasive code modification. This broad, thorough study is the largest in existence and potentially an invaluable guide for future users of similarity detection in source code.

show abstract

“…Source code plagiarism detection [18] is an evolving application area aimed at preventing incidences of software theft at the coding level. A research in this regard by [19] attempts to detect source code plagiarism using an information retrieval technique known as latent semantic analysis by deriving semantic information from source-code files. The application of plagiarism detection as a fundamental part of conference paper review has been proposed by [20].…”

Section: Related Researchmentioning

confidence: 99%

Plagiarism Index Estimation Algorithm: A Quantitative Approach

Eze¹,

Kamboh²

2015

IJCA

View full text Add to dashboard Cite

Plagiarism has remained a serious setback especially in the academia. It is a major source of intellectual theft since it gives credits for scientific innovations to those who do not merit them. A number of efforts have been made by researchers to tackle plagiarism. However, one perceived research gap is the need to evolve verifiable computational techniques for detecting and quantifying the degree of plagiarism in digitized documents. This current research tackles this problem through a specialized plagiarism detection and quantification algorithm. It begins with a bipartitioned search operation known as F-Search. This is followed by a purge operation which excludes the plagiarized sections discovered during the initial pass, thus giving rise to a fresh search space. The resulting search space is passed through a more thorough search operation known as T-Search. At this stage, the algorithm deals with specific plagiarism hiding tricks termed as whitespace flooding. The final output is a statistic known as the Plagiarism Index, which is a numeric value in the range [0, 1] for estimating the degree of plagiarism. The scope of this research covers the text domain. Each experimental dataset is made up of a set of two documents designed in such a way that one is assumed as the original document, while the second as a plagiarized copy. The system is designed and implemented in MATLAB.

show abstract

An Approach to Source-Code Plagiarism Detection and Investigation Using Latent Semantic Analysis

Cited by 144 publications

References 21 publications

An Analysis of the Effect of Mobile Learning on Lebanese Higher Education

An Analysis of the Effect of Mobile Learning on Lebanese Higher Education

A comparison of code similarity analysers

Plagiarism Index Estimation Algorithm: A Quantitative Approach

Contact Info

Product

Resources

About