Similarity Detection Techniques for Academic Source Code Plagiarism and Collusion: A Review

Karnalim, Oscar; Simón,; Chivers, William

doi:10.1109/tale48000.2019.9225953

Cited by 19 publications

(18 citation statements)

References 88 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ottenstein [15] developed one of the earliest techniques for this task, determining the similarity via four software metrics: the numbers of operators, operands, unique operators, and unique operands. Many similarity detection techniques have been introduced since then, relying on either the submitted source code or the creation process [16].…”

Section: Related Workmentioning

confidence: 99%

“…Based on the similarity measurement, techniques relying on the submitted source code can be further classified into three subcategories: attribute-counting-based, structurebased, and hybrid [16].…”

Section: Related Workmentioning

confidence: 99%

“…Last but not least, according to a quick observation of the similarity detection tools listed in a recent literature review [16], a side-by-side comparison module is implemented by only about one fifth of them despite the benefits. This is possibly due to the difficulty of creating such a module from scratch.…”

Section: Related Workmentioning

confidence: 99%

“…Fourth, it uses a string matching algorithm, RKRGST [24], to check for similar code fragments, and sequence alignment [44] to check for similar comments. These are used in preference to other algorithms as they are widely-used [16] and the results are relatively easy to interpret. Fifth, while STRANGE might not recognize all similarities detected by a more advanced tool in which it may be embedded, its output can still be used for initial observation, as it recognizes common variations in code [43], especially those that are trivial but distracting.…”

Section: The Toolmentioning

confidence: 99%

“…However, it is still preferred for reasons of time efficiency; more computation and contextual information are required for higher accuracy. Further, RKRGST's result is arguably acceptable as that algorithm is often used for detecting code similarities [16].…”

Section: Other Modesmentioning

confidence: 99%

See 4 more Smart Citations

Explanation in Code Similarity Investigation

Karnalim

Simón

2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

When using code similarity detection to uncover code plagiarism and collusion, the marker needs to determine whether any detected similarities might be the result of coincidence. But understanding the similarities can be difficult and might be prone to human error, because few tools facilitate the investigation process, and if they do, the similarities are not explicitly explained in human language. This paper presents STRANGE, an investigation module that exclusively explains code similarities in natural language (English and Indonesian). For the purpose of reusability, STRANGE can be embedded in JPlag and other code similarity detection tools. It can also act as a standalone tool for measuring source code similarity. Our evaluation shows that STRANGE is more helpful than JPlag in the investigation process since it explains the similarities in natural language. Further, its effectiveness is comparable to that of JPlag but higher on trivial disguises of the sort that novice students will tend to apply when disguising copied code.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: The Toolmentioning

confidence: 99%

Section: Other Modesmentioning

confidence: 99%

See 3 more Smart Citations

Explanation in Code Similarity Investigation

Karnalim

Simón

2021

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

Layered similarity detection for programming plagiarism and collusion on weekly assessments

Karnalim

Simón²,

Chivers

2022

Comp Applic In Engineering

Self Cite

View full text Add to dashboard Cite

When weekly programming assessments are used, it is often the case that some of them are either trivial or strongly directed. Common code similarity detectors are not particularly helpful with such assessments: some potential instances of misconduct are not selected for manual investigation as all submissions are expected to be similar and it is not feasible to check them all. Several dedicated similarity detectors have been developed to work with such assessments, but the experience is required to determine when to use them. This paper presents a similarity detector that works on many kinds of weekly assessments. It combines three‐layered types of similarity so that even within a set of highly similar submissions, program pairs are still sorted according to their levels of similarity. Our similarity detector is more effective than JPlag in distinguishing similar programs and helping to identify plagiarism and collusion. The similarity detector is slower than JPlag, but the longer execution time is partly offset by some optimization that has no negative impact on the effectiveness. As weekly assessments seldom entail large submissions, the execution time does not appear to be a barrier to use.

show abstract