ケイリョウ　ブンタイロン　シュホウ　ニヨル　エイサクブン　ノ　ゴイ　ヒカク　ゴガク　キョウイク　ヘノ　アプローチ

Stylometric Methods for Lexical Comparison of English Compositions : An Approach to Language Teaching

フジタ, イク

Fujita, Iku

Division of Language and Culture, Graduate School of Humanities, Osaka University

DeepL

機械翻訳

英作文

計量文体論

machine translation

English writing

Stylometry

This paper analyzes three types of written English texts produced by university students in Japan, DeepL, and ChatGPT using stylometric approaches, cluster analysis, and correspondence analysis. The advent of generative AIs is astonishing and machine translators have made our lives easier these days. However, the negative impact on education has also been a concern and discussed. The author of this paper requested university students to write brief essays in English and Japanese about health in their “Data Science” class. The Japanese articles were then translated into English using DeepL. The target corpus comprises English essays written by students, translated by DeepL, and essays generated by ChatGPT. During the class, students analyzed the corpus using the same methods. This paper presents the results of the analyses and the students’ discoveries through text mining. It also discusses how these findings affect their understanding and awareness of the use of generative AIs and machine translators.