A Semantic Approach towards CWM-based ETL Processes

Nowadays, on the basis of a common standard for metadata representation and interchange mechanism in data warehouse environments, Common Warehouse Metamodel (CWM) – based ETL processes still has to face significant challenges in semantically and systematically integrating heterogeneous sources to data warehouse. In this context, we focus on proposing an ontology-based ETL framework for covering schema integration as well as semantic integration. In our approach, beside the schema-based semantics in CWM-compliant metamodels, semantic interoperability in ETL processes can be improved by means of an ontology-based foundation to better representation, and management of the underlying domain semantics. Furthermore, within the scope of this paper, a set of CWM-based modelling constructs driven by ontology for the definition of metadata required for ETL processes is defined, facilitating the extraction, transformation and loading of useful data from distributed and heterogeneous sources. Thus, the role of interconnecting CWM and semantic technologies in populating data warehousing systems with quality data and providing data warehouse an integrated and reconciled view of data is highlighted.