Date of Award
1988
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Computer Science
First Advisor
Peter P. Chen
Abstract
An ERG (Entity-Relationship Graph) can be used to provide a semantic structure to a relational database system. An ERG is defined by local regions. A local region contains two nodes of entity types and a node of relationship type. The semantic constraints of the database represented by the ERG (Entity-Relationship Graph) can be used to enforce the global integrity of the database system. A query is mapped onto the ERG to obtain an ERQG (Entity-Relationship Query Graph). This mapping can be specified by the user by navigating the database or automatically allocated by the system via a universal relation interface. The ERQG representation of a query can be semantically decomposed into a sequence of Local Regions. These Local Regions can then be processed according to their order in the query. The ER-semijoin operation is introduced to process this sequence of Local Regions. Using this approach, architectures of database systems are proposed--two-phase interface and one-phase interface. An implementation of a user interface is also discussed.
Recommended Citation
Chen, Hung-pin, "Query Processing on the Entity-Relationship Graph Based Relational Database Systems." (1988). LSU Historical Dissertations and Theses. 4490.
https://repository.lsu.edu/gradschool_disstheses/4490
Pages
173
DOI
10.31390/gradschool_disstheses.4490