Data abstraction for designing database-intensive applications

Abstract

There is a growing exchange of ideas amongst Artificial Intelligence (AI), Database (DB) and Programming Language (PL) researchers concerning conceptual modelling of complex, object-oriented applications. The complexity of these applications arises from complicated structural and behavioral properties which change through time; concurrent, interactive access by users with different processing needs over a shared database; information locality (i.e., DB views, PL data abstractions, AI perspectives); and primarily update-oriented transactions. Two main problems raised by these applications are: managing the intellectual complexity of their design, development and evolution, and defining and ensuring semantic integrity.