Detecting Back Translation Plagiarism in Scholarly Writing

A Comparative Analysis of Interlanguage Matching

Abstract

Back translation plagiarism has been identified as a new form of plagiarism has become prevalent in scholarly literature. A sophisticated cyber plagiarism makes it difficult for available anti-plagiarism checkers to effectively detect surreptitious copying of content across languages. This paper investigates strategies to ascertain how back translation plagiarism can be tracked through semantic and stylistic checks of corpus generated from back translated text across selected languages. To achieve this objective, we adopt a comparative analysis of extract from selected online scholarly text which was back translated across distant and close language pairs. The study reveals that back translation plagiarism between close language pair shows significant plagiarism issues with lots of identical intelligible content matching as the original while distant language pair shows no case of plagiarism but more unintelligible sentences. Based on the findings, the study recommends the use of sophisticated plagiarism security systems to avert this academic crime in Higher Education. The paper calls for creation of online database of all undergraduate and postgraduate theses and dissertation submitted across the globe. It also suggests the need for further improvements on the quality and capacity of available online plagiarism checkers to be able to detect this form of plagiarism.