We expect pStore to provide a set of default schemata, like the ones above (and possibly more). However, we expect users to modify these schemata. For example, in many data management applications, relationships among data objects are identified after the objects are created and may change during the lifetime of the objects, as their usage changes. The usage of data and metadata is often unpredictable and may depend on the actual user or workload. Incremental elaboration of data object classes and their properties is often inevitable. We also expect users to define their own schemata and share them in ad-hoc manners to cover application or site-specific requirements among communities of users.
RDF supports dynamic evolution of schema in multiple ways. First, it supports refinement of schema through class inheritance and property polymorphism. Second, the namespace feature of RDF allows for schemata to evolve differently in different contexts, such as application versions or user communities. Last, but not least, the fact that RDF provides a machine-readable notation, facilitates the design of programmable interfaces and tools that allow for automatic extraction, manipulation and exchange of relations and schemata.