COUPLING ONTOLOGY WITH EARTH SCIENCE MARKUP LANGUAGE FOR SCIENTIFIC DATASET DESCRIPTION

The large amount of heterogeneity in scientific data formats often leads to limited data usage and exploitation due to the inability of science applications to decipher all the data formats. The Earth Science Markup Language (ESML) is an interchange technology solution for this scientific data formats-application interoperability problem. ESML consists of three components that constitute this interchange technology as follows: ESML Schema; ESML Description File(s); and the ESML Library. The ESML Schema is a specification in XML that defines the grammar/rules for describing the structure of the data file in terms of bits and bytes. The ESML Description File(s) are actual valid instances of structural metadata descriptions for any given data set(s). The ESML Library is the middleware utilized by an application to read and extract data from various formats by deciphering these ESML Description Files. The ESML Schema design specifies the complete structural metadata for the description different scientific data formats. However, for applications to fully exploit the data, the metadata description should also contain semantic information. Because of the scope of many and varied science domains, it is impractical to specify general semantic information within the ESML Schema. A better approach is to allow different science domains to develop their own ontologies and allow the ESML Description Files to include these ontologies. This paper will describe the mechanism to provide complete structural and semantic metadata information about a data set by coupling ontologies within the ESML Description Files. It will also describe a “smart” application prototype that utilizes the structural metadata to read the data file and the semantic metadata to intelligently “use” the data without human intervention.