Constructing a Dataspace Based on Metadata and Ontology for Complicated Scientific Data Management

As data sharing and integration has the properties of data heterogeneity, semantic relationship, dynamicity, and immediate efficiency, dataspace has been proposed as a data management approach for diverse applications. This paper introduces a framework of constructing a dataspace based on metadata and ontology, especially for complicated scientific data management. Firstly, supported by the ontology, this method sets up a metamodel independent from the various platforms of data sources, and then applies the metamodel to describe the participants of the dataspace and the relationship between them, that is producing metadata to describe the dataspace. Based on the unified metadata, the search and query service can be provided for all the data within the dataspace.