Controlled Redundancy in Object Bases: Optimization Methods for Organizational Data Modelling

We review the de nition and then analyse two optimization techniques that were developed for objectoriented databases. Both of these approaches are based on incorporating redundancy into the object base. However, the redundancy is entirely controlled by the system, rather than burdening the user with this task. One technique, called Access Support Relations, is based on redundantly maintaining frequently traversed object paths. The other, Function Materialization, is founded on precomputing particularly performance critical functions. In both cases, the redundantly maintained information can be viewed as an index to speed up object base queries. We discuss the exploitation of these optimization approaches on two sample, yet representative applications from the area of organizational data modelling.