Title of Paper:  Content-Based Metadata System: A Workbench to Prototype Data Mining Concepts


Principal Author: R. Suresh


Abstract: The purpose of this prototype is to identify appropriate level-3 datasets and build a content-metadata "warehouse". This warehouse will serve as a data workbench for other data mining prototypes. Earth science archives hold many level-3 products that contain higher level processed, geophysical parameter data on a space-time grid. These products were derived based on widely accepted algorithms. If brought to an on-line environment, they could serve as content-based metadata for "mining" other data held in NASA archives.  This can offer an environment to explore what is and isn't useful to the science community so that true and useful data mining algorithms can be developed and implemented as part of NASA's data systems.


This prototype currently has analyzed and identified level-3 data products from EOS Terra, TRMM and other missions. A survey of science users was conducted to identify the geophysical parameters of importance. COTS software E-cognition, Database and Mark up language technologies are being investigated as part of this effort. This presentation will discuss the current status of the project and potential technologies that will be used.


This prototype will benefit the community in understanding the requirements for a data mining tool. The metadata repository that will be developed as part of this project serves as a resource for other data mining projects. Once all phases are completed, this may also serve as a very useful research planning tool that does not exist today.