2018 Fiscal Year Final Research Report
Developing high-performance and high-functional wikification using multilinguality of Wikipedia
| Project/Area Number |
15K16096
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Library and information science/Humanistic social informatics
|
|---|
| Research Institution | Shizuoka University |
|---|
Principal Investigator |
|
|---|
| Research Collaborator |
KAJI hiroyuki
KOTANI ryota
MURAKAMI ryosuke
NISHIDA masafumi
NISHIMURA masafumi
|
|---|
| Project Period (FY) |
2015-04-01 – 2019-03-31
|
| Keywords | Wikification / エンティティリンキング / Wikipedia / 多言語 / 情報組織化 |
|---|
| Outline of Final Research Achievements |
We proposed a method for converting links in Wikipedia articles of other languages into Japanese for wikification task, which links important phrases in text to Wikipedia articles. The experimental results did not show improvement of performance for link destination determination, although the number of converted links increased. In addition, for each of subtasks, anchor detection and link destination determination, we proposed methods based on new features and abstractness of co-occurring words, and confirmed improvement of performance.
|
| Free Research Field |
自然言語処理
|
| Academic Significance and Societal Importance of the Research Achievements |
テキストにWikipedia記事へのリンクを付与することにより、テキストの理解可能性を高めるのみならず、情報抽出等の自然言語処理タスクへの応用も期待される。本研究は、従来のWikipediaのリンク言語間変換の手法を発展させてwikificationに応用し、全言語のWikipediaのリンクデータを用いてwikificationの性能向上を図る方法を検討したものであり、リンク数を増加させることで新たなリンクの提案を一定の精度で行うことができる可能性を示した。
|
