Schema integration in heterogeneous distributed data base systems using semantic data modeling
Cyrus Azarbod · 1993
Interoperability among heterogeneous distributed data base systems (HDDBS) continues to be a topic of interest within the fields of computer science, and it is identified as an active area of research for the 1990s. Schema integration is considered to be a major concern in HDDBS research. More recently, semantic data modeling (SDM), which is significantly more powerful than other data models used in databases, has been introduced. Incremental concept formation models of machine learning have also demonstrated their applicability in solving some database problems. The Data Base Schema Integration Model (DBSIM) research presented here has implications within both database and machine learning fields. An SDM was presented as a generic interface to convert the local database schemas into a common abstract data model. The converted local data dictionaries were used to build concept dictionaries. The incremental concept formation algorithm was presented as a viable model for classifying concepts into a hierarchical structure. This research extended to previous models of incremental concept formation by presenting multiple, logically connected concept hierarchies that support similar concepts in their similarity function, aggregate concepts, and weak concepts. A series of algorithms was developed to use the concept dictionary and concept hierarchies and to integrate the local database schemas into a federated data dictionary. DBSIM maintains the unique mapping between each local database schema and the federated schema. The models presented here were compared with related research, and the main differences were identified.