2002 Fiscal Year Final Research Report Summary
Automatic Extraction of Sentence Pattern Pairs from a Large Scale Corpus for Japanese to English
| Project/Area Number |
12558035
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 展開研究 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Tottori University |
|---|
Principal Investigator |
IKEHARA Satoru Tottori University, Faculty of Engineering, Professor, 工学部, 教授 (70283968)
|
|---|
| Project Period (FY) |
2000 – 2002
|
| Keywords | Natural Language Processing / Sentence Pattern / Automatic Knowledge Acquisition / 知識自動獲得 |
|---|
| Research Abstract |
In order to realize the semantic analysis needed for machine translations, a sentence pattern pair extraction program was developed. This program was examined to evaluate the effect how it is able to automatically extract semantic structural sentence pattern pairs from a large scale corpus which contains one million sentence pairs for Japanese to English. In this experiment, first, 150,000 sentence pairs were extracted for complex and compound sentence. And, from the result, 130,000, 100,000 and 20,000 sentence pattern pairs were extracted for word level sentence pattern, phrase level sentence pattern and clause level sentence pattern. From these experiment showed the high performance for half automatic sentence structure extraction. Thus it came possible to make a plan to develop a large scale semantic knowledge base of the sentence structures for Japanese and English. Based on this result, the new project to develop such a knowledge base was started this year.
|
