81006682

open access

Version of Record

Data Collection and Annotation of Relatively Spontaneous and Relatively Extended Elicited Utterances by English Learners in Undergraduate Japanese Courses

English (英語)

Learner Corpus Studies in Asia and the World

1

179-197

神戸大学国際コミュニケーションセンター

2013-03-23

2014-06-24

In this paper we report on an eight-year project for collecting spoken utterances by English learners enrolled in Japanese university English language classes. Most of our efforts so far have focused on data collection, with some rudimentary manual transcriptions of a very limited part of the collected utterances. The raw audio and video recordings of students interacting in English and Japanese come to a total of 200 hours or so per semester, as seven to eleven groups of students engage in the activity for 25 to 30 minutes with a 90-minute class that convene 15 times in a semester. The core data corresponds roughly to one third or one fifth of these raw materials. The portion manually transcribed so far represents very roughly about 1/20 of the core data to be compiled and the transcription came to 150,000 words including 10% of filler-like expressions. This suggests that when we transcribe the entire core data among the collected spoken utterances, it will come to something like 3,000,000 words. In this presentation, we describe our data collection and audio transcription interface and then touch upon potentials for new research that can be conducted on the basis of the collected data. We propose a new way to look at some of the natures of spontaneously spoken English by Japanese learners in interactive settings.

紀要論文