| Budget Amount *help |
¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2014: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2013: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2012: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
|
|---|
| Outline of Final Research Achievements |
The research project developed the method to optimize the training data set of domain adaptation for word sense disambiguation. Firstly a number of training data set were generated and then the best one is selected using some criteria. The degree of confidence, which is a criterion used for the instance selection when active learning is carried out, and a score using LOO-Bound, which is an error rate of SVM when leave-one-out estimation is performed, are used for the criterion.
|
