| Project/Area Number |
02680035
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Informatics
|
|---|
| Research Institution | Fukuoka Institute of Technology |
|---|
Principal Investigator |
TAMACHI Tsuneo Fukuoka Institute of Technology, Faculty of Engineering, Professor, 工学部, 教授 (90037656)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TERADA Hideo Fukuoka Institute of Technology, Faculty of Engineering, Lecturer, 工学部, 講師 (20148901)
|
|---|
| Project Period (FY) |
1990 – 1991
|
| Project Status |
Completed (Fiscal Year 1991)
|
| Budget Amount *help |
¥1,400,000 (Direct Cost: ¥1,400,000)
Fiscal Year 1991: ¥300,000 (Direct Cost: ¥300,000)
Fiscal Year 1990: ¥1,100,000 (Direct Cost: ¥1,100,000)
|
|---|
| Keywords | Natural Language Processing / Machine Translation / Chinese |
|---|
| Research Abstract |
We have engaged in the work of the improvement of Chinese-Japanese machine translation system for these 2 years. The outlines of research results are as follows :
1. As Chinese sentences are written without inserting spaces between words, input sentences must be first divided into words. We constructed a segmentation program, which performs a morphological analysis and divide a continuous Chinese sentence into words. We use this program before translation to get a segmented text in the case of batch translation.
2. We analyzed the usage of complements, which are very often used in Chinese, and examined the way of processing them.
3. We extended the expression of semantic marker from 1 character into 2 characters to improve the selection of suitable word meaning from an ambiguous word.
4. Our system is intended to be utilized in translating documents primarily in scientific add technical field. We tried translation experiments using a scientific article about computers extracted from Chinese newspaper to enrich the items of the word dictionary and the syntax dictionary.
5. As scientific documents use many technical terms and compound words, we must examine how to process technical terms. We did preliminary study on the processing of technical terms.
6. We improved the system to have backtrack mechanism, so that we can get more than one translation results when a word has several meanings.
7. We constructed an English discourse understanding system named IMAGES-II. This system translates input discourses into semantic structures and answers questions. The knowledge representations and the processing algorithms employed in this system are free from any natural language.
|
