A semantic framework for heterogeneous database environments

Describes a semantic framework for the analysis of global view update operations in a heterogeneous database environment. A semantic data model is used to form global conceptual schemas and to represent the meta data of global schemas using a self-describing approach. The semantic model is also used to represent the meta data of local schemas and the mapping between local and global schemas. Set-oriented views of objects are then formed over the global conceptual schema using an object algebra. An approach known as constraint analysis is applied to the analysis of view update operations, identifying side effects and alternative update translations. The constraint analysis process helps to refine the scope of a global view based on the constraints of the underlying local schemas and the types of update operations to be performed. The constraint analysis process also supports semantic query optimization. The approach described in this paper provides a uniform interface for defining global object-oriented views, together with analysis tools.<>