A framework for data base semantic integrity

A structured framework is provided for describing the semantic integrity requirements of a data base. The semantic integrity of a data base is said to be violated when the data base ceases to represent a legitimate configuration of the application domain it is intended to model. In the context of the relational data model, it is possible to identify multiple levels of semantic integrity information. Relation constraints comprise one such level; they express the semantic information not contained in the structure of the relations nor in the identity of their underlying domains. The three components of a relation constraint are considered: (1) the assertion (a predicate on the state of the data base or on transitions between data base states), (2) the validity requirement (the occasion(s) at which the assertion must hold), and (3) the violation-action (the action that is to occur if the assertion does not hold at a time when it should). A framework for relation constraints is presented. Details of a structured classification scheme are outlined. This scheme is intended to form a basis for a high level, well-directed, and disciplined methodology for the design of relational data bases. Emphasis is placed on the assertion component of relation constraints. Assertions are viewed as more than expressions of some relationship among different values in a data base; assertions single out the data that is constrained, and state the properties that this data must possess. A classification is provided of the various predicate types used to identify constrained data and to state the properties which they are to possess. Approaches to the relation constraint assertion specification process are discussed.