81011993

open access

Version of Record

Developing Semantic-based DDL Based on a Comparative Study of the Verb Use of British and Japanese Students

Learner Corpus Studies in Asia and the World

4

41-66

神戸大学国際コミュニケーションセンター

2020-03-17

2020-03-13

This study aims to develop semantic-based data-driven learning (DDL) for use in Japanese secondary school classrooms by comparing the use of verbs by British elementary and lower secondary school students and essays written by Japanese secondary school students. Three specialized corpora were used: entries from the BBC’s 500 Words competition written by 5- to 9-year old (BBC_5–9) and 10- to 13-year old (BBC_10–13) British students from 2014 through 2017, and the learner corpus (LC_13–15) of Japanese secondary school students from 2006 through 2008. The collected data were collated as annotated corpora adopting both the POS (part of speech) and semantic tags. The study was designed to address three points: identifying the quantitative differences in verb use in each corpus, examining the qualitative difference of these verbs by correspondence analyses based on annotated POS and semantic tags, and developing DDL materials. The study compared the Type-Token Ratio of verb use and found that older British students use a wider variety of verbs than Japanese secondary school and younger British students. Furthermore, this study identified 17 characteristic verbs of the BBC_5–9 and BBC_10–13 corpora classified by semantic categories in the 100 most frequently used verbs. Based on the results, this study suggests four different types of DDL for a learner-centered approach to discovery learning to widen learners’ knowledge of verb use by learning the different verbs classified within the same semantic categories through semantic-based DDL.

DDL

紀要論文