Index Data Structures in Object-Oriented Databases

Object-oriented database management systems (OODBMS) are used to imple- ment and maintain large object databases on persistent storage. Regardless whether the underlying database model follows the object-oriented, the rela- tional or the object-relational paradigm, a key feature of any DBMS product is content based access to data sets. On the one hand this feature provides user-friendly query interfaces based on predicates to describe the desired data. On the other hand it poses challenging questions regarding DBMS design and implementation as well as the application development process on top of the DBMS. The reason for the latter is that the actual query performance depends on a technically meaningful use of access support mechanisms. In particular, if chosen and applied properly, such a mechanism speeds up the execution of predicate based queries. In the object-oriented world, such queries may involve arbitrarily complex terms referring to inheritance hierarchies and aggregation paths. These features are attractive at the application level, however, they increase the complexity of appropriate access support mechanisms which are known to be technically non-trivial in the relational world.