| Budget Amount *help |
¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2013: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
Fiscal Year 2012: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
Fiscal Year 2011: ¥1,690,000 (Direct Cost: ¥1,300,000、Indirect Cost: ¥390,000)
|
|---|
| Research Abstract |
We aim to develop a method to verbalize multiple time-series data with words. We calculate the correlation coefficient between two time-series data and convert the data into symbols by SAX. The edit distance of the symbols of two time-series data is calculated and expressed with words. We also propose a method to verbalize the interaction between a human and an object. SAX with dynamic programming is applied to the time-series data to extract their patterns, and then the correspondence between the patterns and their semantic labels are learned by a log-linear model. We build a bi-gram model based on the collected linguistic resources. The most likely sentences to explain the interaction are generated by solving the bi-gram model with dynamic programming. Moreover, to enhance the linguistic resources used to generate sentences, we adopt transfer learning of n-gram language models. Through experiments, we have confirmed that our proposed framework works well.
|
