ITR/IM+AP: A Conceptual Study of Multi-Nested Self-Populating Scientific Data Centers

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

This is an Information Technology Research proposal to develop a new concept of "data-crawling" software framework allowing a freelance "sailing" across various scientific datasets available on the Web. The existing World Data Center (WDC) System serves the worldwide scientific community in providing free access to a number of global geophysical parameters. However, this system still requires the (mandatory or voluntary) submission of locally collected data to the data centers, housed at certain institutions or organizations at different geographic locations and mainly organized by the corresponding scientific discipline. At the same time, a huge number of scientific digital databases are currently placed on the World Wide Web without submission to any data center, fulfilling the WDC System's goal of free access to the data, but (most importantly) promoting recognition of the local institution efforts. This creates a need for sophisticated methods to search and to recognize the data type, and then to retrieve a certain amount of data from these databases. The "data-crawler" to be developed will check the data types and levels their accessibility in establishing mutually agreed interconnections between a data host and a requesting "data center" and then supplying retrieved data samples (subsets) to this single- or multi-nested, self-populating data center. The research project will yield a working prototype of the "data-crawler" search engine with practical applications to the Web-based scientific datasets. This will allow users to improve their capabilities in environmental informatics and geoscience modeling and representation, as well as facilitate scientific research via the support of ubiquitous content infrastructure. The latter will help in organizing a "Virtual Global Magnetic Observatory" (VGMO) where the worldwide geomagnetic data can be visualized on a global scale through the feed from multi-nested data source sites. The effort will focus on the disciplines of solar-terrestrial physics and geophysics, but the proposed software framework can be applied to any scientific (or even commercial) databases distributed without a pre-set structure over the Web.