Japanese
| Documents are explanatory structure of events in them, and they should be organized in a hierarchy under subsumption relationships between them. From this viewpoint, a bottom-up data mining algorithm that enumerates minimally generalized event sequences from a family of two or more event sequences is presented. A generalized event sequence thus obtained by this algorithm is regarded as an abstract story description. However, such a naive miner never examines a global structure of documents. Understanding the global structure is a key to find an informative abstract description explaining the documents properly.For this purpose, we propose a text summarization system taking not only major parts of documents but also some contextual parts connecting the major parts into account. This enables us to find an informative abstract common description between two or more documents. |
|
We perform various behaviors as parts of our everyday lives. In scientific fields such as quantum theory and cosmology, gstandard modelsh exist to explain and generate most phenomena. However, nothing yet exists which might represent a standard model of everyday life. In order to create such a model of daily activity, both the sensing technology needed to define phenomena and the modeling technology needed to construct a logical system based in some quantification of this activity are essential. Recent years have seen the introduction of technologies for sensing physical phenomena in total space using ubiquitous sensor technology, technologies for sensing worldwide social phenomena using Internet technology, and also the use of knowledge acquisition technology based on the large-scale data gained using these sensing technologies. This implies that everyday life is becoming a new target of science and technology. From knowledge acquisition and circulation of everyday life, in this exhibition, we present several real world applications on the following titles. |
|
 |
