The Implementation of Using Medical Ontologies in Plagiarism Detection

Abstract

This paper aims to present a new algorithm implemented for detecting plagiarism using semantic web tools and notions. For increasing detection accuracy domain ontology could be used in addition to global semantic resources. Using global semantic resources increases the effect of ambiguity therefore disambiguation technique was used. Not all semantically similar texts are plagiarized. So, other detection techniques were used to reduce false positive results. Given that our work has been implemented in medical disciplines and for texts written in English, it presents a generic algorithm that can be adapted for different disciplines and languages. For medical discipline, a set of medical Ontologies were used for enriching extracted medical terms. In addition, WordNet was used for enriching global terms. The test results of the algorithm shows that it was able to detect advanced types of plagiarism that are out of the reach of classical methods such as: using word synonyms, word re-ordering, text re-styling and other natural languages techniques which are usually used to hide the plagiarism action.