A semantic data model approach to logical data independence

Abstract

Semantic data models were originally introduced as a tool for data description and design. SDMI (Semantic Data Model Interface) investigates their use in another role - as the basis for a data manipulation language - and shows how this can provide a means of achieving complete logical data independence. With SDMI users view a database simply in terms of objects and their attributes, and can access any combination of related information just by stating the relationship(s) of interest. This paper outlines the problem of logical data dependence that is inherent in all existing database systems and shows how the universal relation approach fails to provide a complete solution. SDMI is described and shown to be a more promising alternative.