Research On Incorporating Semantic Relationships Into OODBs

Part-whole relationship is the familiar relationship between objects in OODBs. In order to describe the model formally. M. Halper has organized the Part-Whole semantic model into four characteristic dimensions: exclusiveness. multiplicity, dependency, and inheritance. Aiming at the disadvantage of this model in expressing the integrity of objects, in this dissertation, we organize the Part-Whole semantic model into five characteristic dimensions: exclusiveness. integrity, multiplicity, dependency, and inheritance, in which the characteristic dimension 'integrity' is to describe the integrity status of objects, and then the integral definition of our model is presented. Our model has enhanced the ability to express the integrity of objects.Part-Whole semantic relationship plays an important role in modeling information models for complex application systems. This is true whether one employs traditional data modeling techniques like ER. or object-oriented data modeling techniques like OMT. UML and Cord/Yourdon. Part-Whole semantic relationship is the familiar relationship between objects, but the popular commercial OODBs do not provide the built-in support for it. So. M. Halper had presented a methodology based on the metaclass mechanism to incorporate the Part-Whole semantic relationships into OODBs. Because there are a few OODBs which support the metaclass mechanism, this methodology has limitation to some degree. Aiming at this problem, in this dissertation, we present a new methodology to incorporate Part-Whole semantic relationship into OODBs. Our methodology is universal for there is no stringent requirement of OODBs.