Data centric science for information society

Abstract

Due to rapid development of information and communication technologies, the methodology of scientific research and the society itself are changing. The present grand challenge is the development of the cyber-enabled methodology for scientific researches to create knowledge based on large scale massive data. To realize this, it is necessary to develop a method of integrating various types of information. Thus the Bayes modeling becomes the key technology. In the latter half of the paper, we focus on time series and present general state-space model and related recursive filtering algorithms. Several examples are presented to show the usefulness of the general state-space model. © 2010 Springer-Verlag Tokyo.