Automatic Plagiarism Detection Using Word-Sentence Based S-gram
Pakinee Aimmanee* Author for corresponding; e-mail address: pakinee@siit.tu.ac.th
Volume: Vol.38 (SPECIAL ISSUE 2011)
Research Article
DOI:
Received: 15 July 2010, Revised: -, Accepted: 31 January 2011, Published: -
Citation: Aimmanee P., Automatic Plagiarism Detection Using Word-Sentence Based S-gram, Chiang Mai Journal of Science, 2011; 38(ECIAL): 1-7.
Abstract
Plagiarism is an academic problem that is caught more and more each year. Common tricks that the cheaters normally use is inserting and removing a few extra terms, sentences, or paragraph to the original copy to trick the reader that the plagiarist copy and the original copy are unalike. This paper provides a new way to detect the plagiarism by checking the similarity between sentences, and paragraphs using s-grams which is an n-gram unit that allows some terms to be skipped to be used in term, sentence, and paragraph representation in finding the similarity between two documents. We discovered that our technique considerably outperforms the existing techniques. Keywords: s-gram, sentence similarity, paragraph similarity, document similarity, and plagiarism.