| Project/Area Number |
17K00299
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | University of Yamanashi |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-04-01 – 2022-03-31
|
| Project Status |
Completed (Fiscal Year 2021)
|
| Budget Amount *help |
¥4,550,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥1,050,000)
Fiscal Year 2019: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2018: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2017: ¥2,080,000 (Direct Cost: ¥1,600,000、Indirect Cost: ¥480,000)
|
|---|
| Keywords | 分野依存語義 / 文書分類 / 語義 / パラフレーズ同定 / CNNs / 転移学習 / 分野の階層構造 / マルチタスク学習 / 分野語義 / 深層学習 / ヒトの修正過程 |
|---|
| Outline of Final Research Achievements |
The objective of this research is to develop lexical-semantic processing and supervised deep learning technique for text classification that classify test documents which are different period of the training documents. We have published five papers as our research output this pear. More specifically, (1) identification of domain-specific senses, (2) two approaches for word sense disambiguation, and (3) semantic equivalence. Furthermore, we proposed a method to utilize the lexical semantics of a label assigned to the training data and semantics of words in a document to improve the overall performance of text classification task. In future, we are going to extend our current domain-specific senses to fewl labeled training instances.
|
| Academic Significance and Societal Importance of the Research Achievements |
本研究で提案する分野語義獲得手法は, 人手で作成した辞書とは異なり辞書の改版や異なる体系を持つ辞書, 異なる言語辞書に対しても柔軟に適用可能な枠組みであり, このことは, インターネットの普及に伴う情報の量と質に十分対応可能な言語処理手法の一つを提案することにも繋がる. 開発した分野語義データベース, 及び学習手法は,情報検索をはじめ, 質問応答やフィルタリングなど文書分類以外の様々なタスクに適用可能である. 併せて, 分野語義データベースの利用過程で生じる問題点は, 現在の学習理論, 言語処理技術, 及び言語資源にも還元することができることから, 本研究の学術的意義は極めて大きい．
|
