Logical Data Independence Reconsidered (Extended Abstract)

Logical data independence is central to a number of database research problems including data integration and keyword search. In this paper, a data model that unifies several of the most widely adopted data models is studied. The key is to disassociate metadata from particular roles. A high-level, context-based (or semantically-based) query language is introduced, and applications to the aforementioned areas of research are demonstrated briefly.