Handling uncertainty and exceptions with defeasible dependencies

Recently, the merger of artificial intelligence tech­ niques with database management concepts has sparked much interest. One important area is the development of a theory for relational database design based on absolute, incomplete, and uncertain knowledge about the domain. Tire approach presented here brings a new dimension to the traditional theory of relational database design based solely on absolute knowledge. Specifically, we present the foundations of a theory of defeasible dependencies. Defeasible Dependencies are closely related to functional dependencies, but they allow exceptions as well as uncertain relationships between groups of attri­ butes to be represented. That is, functional dependencies rely on the fact that dependent relationships among the attributes of interest arc known for certain. Defeasible dependencies provide a capability for handling these rela­ tionships even when exceptions exist. In addition, defeasible dependencies can represent relationships that the designer would not specify as functional because of some degree of uncertainty regarding the relationship. This allows designers to develop database applications in a more natural and flexible manner than currently possi­ ble.