2009 Fiscal Year Final Research Report
Structurized Term Extraction from Academic Text Corpora
| Project/Area Number |
19500135
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | National Institute of Informatics |
|---|
Principal Investigator |
KOYAMA Teruo National Institute of Informatics, 情報社会相関研究系, 教授 (80124410)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TAKEUCHI Koichi 岡山大学, 大学院・自然科学研究科, 講師 (80311174)
|
|---|
| Project Period (FY) |
2007 – 2009
|
| Keywords | 用語抽出 / 用語体系化 / 自然言語処理 / 形態素解析 |
|---|
| Research Abstract |
In this study, we established a method for comprehensive term extraction from domain text corpora with high precision. The method is based on basically two new principles. One is the reconsideration and modification of Japanese morpheme classification, and another is the evaluating the certainty of composite boarders. We also have developed methods to structurize terms from two points of view, namely, nesting relations between composites, and the term relationships to various research subdomains.
|
