A Framework Model Study for Ontology-Driven ETL Processes

ETL is a critical task performed at the early stages of a DW project. To solve heterogeneity problems of different data sources in the processes, this paper proposes a framework model for ontology-driven ETL processes by using semantic Web technologies and discusses how ontologies are used to support the data integration. The framework consists of four main phases: metadata abstract, ontology map, rule reason and ETL direction. The main ideas is to abstract local ontologies from metadata of different data sources, global ontology from data warehouse via OWL DL under the direction of domain ontology which is built by domain experts first, next map these ontologies to find the connotative semantic relations, then reason these relations to get exact ETL rules which can be used to direct the ETL processes and assist the performing of DW project.