Chaining Distributed Services for Mining, Fusion, and Visualization

The Information Technology and Systems Center (ITSC) at the University of Alabama in Huntsville, as a NASA Earth Science Information Partner (ESIP), is researching the use of distributed services for data mining, subsetting, image processing, image/map generation and other spatially oriented data applications. Many of these applications lend themselves well to being broken down into discrete processes that can be combined in different ways to solve more complex problems. A service chain may combine a variety of processing or analysis services, like data mining, with standard protocols for returning results to the user, such as the Web Mapping Services defined by the OpenGIS Consortium. However, applying distributed service technologies to the processing of Earth Science data sets presents unique problems. Conventional distributed service protocols such as the Simple Object Access Protocol (SOAP) tend to be best suited for exchanging textual data or relatively small amounts of binary data, but processing Earth Science data typically means working with files on the order of megabytes or even gigabytes in size. ITSC researchers have been investigating alternative solutions for providing distributed services that are capable of processing and passing large amounts of data in a service chaining architecture. ITSC will demonstrate integrating simple, autonomous data processing services into chains that efficiently process large Earth Science data sets and can be easily reused in other distributed applications. This presentation will illustrate how the Earth Science Markup Language (ESML) can be used in this environment to describe the output format of a service in the same way it is used to describe the format of a data set.