Improving Plagiarism Detection Using Genetic Algorithm

Detecting instances of plagiarism in student home-work, with program code in particular, is a subject of active research for over 30 years. One of the early proposed methods was extraction and comparison of source-code metrics. Even though this approach has low algorithmic complexity, it is rarely u...

Full description

Saved in:
Bibliographic Details
Published in2019 42nd International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO) pp. 571 - 576
Main Authors Pajic, Enil, Ljubovic, Vedran
Format Conference Proceeding
LanguageEnglish
Published Croatian Society MIPRO 01.05.2019
Subjects
feature extraction
plagiarism detection
real-coded genetic algorithm
Online AccessGet full text

Cover

More Information
Summary:Detecting instances of plagiarism in student home-work, with program code in particular, is a subject of active research for over 30 years. One of the early proposed methods was extraction and comparison of source-code metrics. Even though this approach has low algorithmic complexity, it is rarely used in recent papers with some authors claiming that better results are obtained using other methods. In this paper, plagiarism detection is treated as an information retrieval problem, specifically query-by-example (QbE). A feature vector is constructed from source metrics and compared using common similarity measures. Further, evolutionary computation methods are used to optimize the similarity measure. It is shown that, by several metrics used, detection results are on par with state-of-the-art methods with significantly lower execution time.
ISSN:2623-8764
DOI:10.23919/MIPRO.2019.8756744