Constraint acquisition you can chase but you cannot find

  • Authors:

  • Sven Hartmann;Sebastian Link;Thu Trinh

  • Affiliations:

  • Massey University, Palmerston North, New Zealand;Massey University, Palmerston North, New Zealand;Massey University, Palmerston North, New Zealand

  • Venue:
  • APCCM '08 Proceedings of the fifth Asia-Pacific conference on Conceptual Modelling - Volume 79
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

We identify established tableaux techniques as an invaluable tool for semantic knowledge acquisition in the design process of relational databases. Sample databases allow users and designers to judge, justify, convey and test their understanding of the semantics of the future database. In the case of integrity constraints such sample data can provide considerable assistance for deciding whether a constraint captures desirable information about the database or not. Since constraints can be particularly difficult to grasp in practice sample databases offer a convenient tool to confirm or reject the usefulness of potential candidate constraints. We pinpoint the Chase and analytical tableau as two tableaux techniques that are able to automatically generate sample databases for large classes of integrity constraints. The Chase can be used for generating sample data that allows us to reject candidate constraints. However, analytical tableaux enable us to find all minimal sample databases which enable us to either accept or reject a candidate constraint.