A context-based schema integration process applied to healthcare data sources

Abstract

Decision-making in healthcare mainly depends on the integration of distributed health information contained in multiple, autonomous and heterogeneous data sources. Many information integration systems have been proposed as a way of offering a concise and uniform view of the distributed data, abstracting out their syntactic, structural and semantic diversities. This paper proposes a context-based schema integration process for a mediator-based information integration system applied to healthcare data sources. The distinguishing feature of this process is to explore and model the contextual information needed for schema-level sense disambiguation in its different steps. This process tackles the semantics of data source schema elements by identifying their meanings and, thereafter, establishing semantic affinities among them. The contextual information is modeled using an ontology-based approach enabling reasoning, reusability and sharing of information. © 2010 Springer-Verlag Berlin Heidelberg.