Contextual information and semi-supervised models for transforming sentence to logical form representation
| Project/Area Number |
22700139
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | Japan Advanced Institute of Science and Technology |
|---|
Principal Investigator |
NGUYEN Minhle 北陸先端科学技術大学院大学, 情報科学研究科, 助教 (30509401)
|
|---|
| Project Period (FY) |
2010 – 2011
|
| Project Status |
Completed (Fiscal Year 2011)
|
| Budget Amount *help |
¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000)
Fiscal Year 2011: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2010: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
|
|---|
| Keywords | 意味解析 / 言語理解 |
|---|
| Research Abstract |
The main goal of our research is to implement the method of using contextual features for improving semantic parsing problems. We also study how unlabeled data could help to implement semantic parsing model further. As a result, we exploited word-cluster models to model a large un-annotated corpus, to extract features for discriminative learning models. In addition, we also introduce a novel semi supervised learning model for semantic parsing with ambiguous supervision. We applied the forest-to-string method for learning the synchronous model between semantic representation and natural language sentence. We also present a novel two-phase framework to learn logical structures of paragraphs in legal articles using machine learning and integer linear programming.
|
Report
(3 results)
Research Products
(28 results)
