Abstract:
SQL designs result from methodologies such as UML or Entity-Relationship
models, description logics, or relational normalization. Independently of the methodology, sample data is promoted by academia and industry to visualize and consolidate the designs produced. SQL table definitions are a standard-compliant encoding of their designers' perception about the semantics of an application domain.
Armstrong sample data visualize these perceptions. We present a tool that computes Armstrong samples for different classes of SQL constraints. Exploiting our tool, we then investigate empirically how these Armstrong samples help design
teams recognize domain semantics. New measures empower us to compute the
distance between constraint sets in order to evaluate the usefulness of our tool.
Extensive experiments con rm that users of our tool are likely to recognize domain
semantics they would overlook otherwise. The tool thereby e ffectively complements existing design methodologies in finding quality schemata that process data efficiently.