1. Project name: Language Engineering and Educational Engineering
2. Outline of the Project:
The purpose of language and educational engineering is to realize advancement of language education by applying outcome of analysis of linguistic performance using an applied linguistics approach with information technology. We have made a large corpus which was obtained from our field work and from various language educational institutes, and have analyzed the corpus with information technology. The outcomes of the analysis of the data have been applied to improve language education.
In an effort of attaining the educational purpose, language and educational engineering has played a role as a means to master methodology and practice of language pedagogy that uses corpora derived from spoken and written languages in practical educational fields. We have actually experienced the role through development of a multi-language corpus, e-learning materials and its assessment. In fact, many new young researchers of linguistics and language pedagogy have acquired familiarity with basis of information technology in this experience.
In an attempt of attaining the research purpose, language and educational engineering have pursued applied research on more advanced language educational methods and assessment techniques of linguistic competency, taking in a combined approach of field linguistics and corpus linguistics with support of a state-of-the-art linguistic information engineering methodology, which has been established in the 21st century COE program. This attempt is not limited to theoretical studies on modeling and generalization of linguistic performance. It contributes truly to forming the center of language educational studies to share profits of its outcomes, as well as expanding studies on linguistics and language pedagogy based on various practical linguistic performances.
3. Core Member in Charge: Sano, Hiroshi (GSACS)
4. Collaborating Researchers:
Kawaguchi, Yuji (GSACS)
Mochizuki, Hajime (GSACS)
5. Progress:
Computerization of the textbook text and tagging of the computerized text was performed. With all textbooks as the parent set, the compositional ratio of respective subject fields was examined. For example, regarding language textbooks, a combined total of 85 junior high and high school textbooks were sold in 2007, of which 31 (36.5%) were able to be recorded as the corpus. The categorical distribution of the 160 corpusized textbooks is shown in the diagram.
Diagram: Breakdown of the number of textbooks, classified by subject
Subject Field |
No. of Sales |
No. Recorded |
Proportion |
Chemistry |
24 |
6 |
25.0% |
Home Economics |
32 |
19 |
59.4% |
Language |
85 |
31 |
36.5% |
Modern Society |
28 |
14 |
50.0% |
Information |
43 |
10 |
23.3% |
Government/ Economics |
17 |
10 |
58.8% |
Biology |
23 |
3 |
13.0% |
Geoscience |
7 |
0 |
0.0% |
Geography |
23 |
13 |
56.5% |
Physics |
17 |
2 |
11.8% |
Health & Physical Education |
10 |
2 |
20.0% |
General Science |
53 |
1 |
1.9% |
Ethics |
11 |
7 |
63.6% |
History |
58 |
39 |
67.2% |
Specialized Industrial Subjects |
218 |
3 |
1.4% |
Total |
649 |
160 |
24.7% |
It may be advisable to decide on the number of textbooks using a certain percentage from each respective subject field, so as to avoid systematic error. The recording ratio shows that there are large dispersions between subject fields, and that there are some fields, such as geoscience, that were not recorded at all. In the future, in order to expand the corpus data, it may be profitable to take the constitution of subjects into consideration, and ultimately to consider a more equal proportion of recorded objects from each respective subject field.
An examination of sentence patterns for Japanese language teaching was conducted. Conversion into retrieval-style format is currently being undertaken in order to extract practical examples from the textbooks.
6. Accomplishments:
A morpheme-tagged corpus was produced. Data was recorded for 160 textbooks (1,400 textbooks were purchased). These consisted of 11 junior high school textbooks and 149 high school textbooks.