Scalable Exploratory Data Mining of Distributed Geoscientific Data

Geoscience studies produce data from various observations , experiments, and simulations at an enormous rate. Exploratory data mining extracts \content information" from massive geoscientiic datasets to extract knowledge and provide a compact summary of the dataset. In this paper, we discuss how database query processing and distributed object management techniques can be used to facilitate geoscientiic data mining and analysis. Some special requirements of large scale geoscientiic data mining that are addressed include geoscientiic data modeling, parallel query processing, and heterogeneous distributed data access.