Uniform and high-level intelligent access to heterogeneous information sources

The goal of this research is to develop automated query formulation capabilities that can be used to generate queries over heterogeneous data sources. Data source ontologies are used as a means to heuristically construct meaningful queries within a reasonable amount of time. To facilitate distributed access and to simplify the development of intelligent interfaces, conceptual-level requests to the query formulator need only specify data items to be retrieved, their roles with respect to each other, and associated constraints. Methods for reasoning over conceptual ontologies and schemas that describe the data sources are then used to infer corresponding queries. By maintaining a conceptual-level perspective of the underlying systems, users and applications are insulated from the ontology structures of underlying data sources and their corresponding query languages. Results from this research will benefit both systems developers and designers seeking to access complex distributed and heterogeneous data in a semantically reasonable and efficient manner.