Paper short abstract:

This paper proposes a multilayered approach for detecting academic misconduct in the Japanese language based on the test results of text-matching detection systems (a.k.a. plagiarism detection tools) performance in the Japanese language.

Paper long abstract:

Digitalization of learning environments in language teaching has been making online and digital materials preferable educational resources for both students and teachers, which contributes to Japanese language teaching (JLT), however, it also increases the possibility of academic misconduct (AM), particularly plagiarism. Taking measures (detective, reactive, proactive) to mitigate AM, the text-matching (a.k.a. plagiarism detection) tools are one of the important elements. However, while discussions about plagiarism in languages and language teaching have gone beyond the lingua franca, studies on the Japanese language (JL) and learners (JLLs) are still scarce, and besides, no study addressed the performance of such tools in JL. Considering the worldwide numbers of JLLs and the digitalization in JLT, it is essential, like all languages, to develop a solid methodology with a pedagogical approach for detecting and preventing AM in JLT. Accordingly, this paper proposes to explore to what extent text-matching tools catch similarities in Japanese texts and respond to the needs of JLLs and educators, besides, providing a solid AM detecting methodology with a pedagogical approach. Analyzing the tools' performance, intentionally-plagiarized Japanese-written documents were compiled from four different online and non-online sources (e.g., Wikipedia). The original text from each source was reconstructed according to the 5 most common plagiarism techniques (e.g., copy&paste, paraphrasing) that students possibly use in assignments. In total, 20 intentionally-plagiarized Japanese texts were tested in 10 text-matching tools. Results show us that tools were found to be relatively inadequate in detecting plagiarism in texts due to the characteristics of the Japanese. Besides, concrete evidence was provided on potential plagiarism techniques that come directly from the structure of the JL. Based on the data 1) a multi-term and multi-layered roadmap for JLT stakeholders to establish a solid foundation for detecting and preventing AM, 2) a concrete and ongoing pedagogical model for JL teacher training institutions to build up equipped JL teachers on AM issues were provided. Researchers believe that the findings presented in this study will be a milestone in proposing a discussion framework on AM not only in JL but also in all ideographic languages.

Keywords: academic honesty, Japanese language, plagiarism, text-matching tools