This preview shows page 1. Sign up to view the full content.
Unformatted text preview: SEMANTICS OF BEHAVIORAL INHERITANCE IN DEDUCTIVE OBJECT-ORIENTED DATABASES Hasan M. Jamil A thesis in The Department of Computer Science Presented in Partial Fulfillment of the Requirements For the Degree of Doctor of Philosophy Concordia University Montreal, Quebec, Canada November 1996 c Hasan M. Jamil, 1996 Concordia University
School of Graduate Studies
This is to certify that the thesis prepared By: Entitled: and submitted in partial ful llment of the requirements for the degree of Mr. Hasan M. Jamil Semantics of Behavioral Inheritance in Deductive Object-Oriented Databases Doctor of Philosophy (Computer Science) complies with the regulations of this University and meets the accepted standards with respect to originality and quality. Signed by the nal examining commitee: Chair External Examiner Examiner Examiner Examiner Supervisor Supervisor Approved Chair of Department or Graduate Program Director 19 Dean of Faculty Semantics of Behavioral Inheritance in Deductive Object-Oriented Databases
Hasan M. Jamil, Ph.D. Concordia University, 1996 We argue that powerful models for supporting next generation database and knowledgebase applications can be built by extending semantic data models in the direction of Object Oriented modeling. It is clear that modeling such new applications will require concepts like modularity, behavioral abstraction, derived schema components, and database knowledge (e.g. constraints). Most of the required concepts are already present in these two seemingly parallel models in the form of the notions like object identity, inheritance, encapsulation, methods, virtual objects and classes. but have important di erences. We present a conceptual Abstract Data Model called the Object Relationship (OR) model which reconciles semantic and object-oriented data models and extends the modeling capability with additional concepts like withdrawal, strati ed constraints, methods in relationships, and encapsulation. We then propose a novel semantics for object-oriented deductive databases as a formalization of the OR model in the direction of F-logic to logically account for behavioral inheritance, con ict resolution in multiple inheritance hierarchies, and overriding. We introduce the ideas of withdrawal, locality, and inheritability of properties (i.e., methods and signatures). Exploiting these ideas, we develop a declarative semantics of behavioral inheritance and overriding without having to resort to non-monotonic reasoning. Con ict resolution in our framework can be achieved both via speci cation and by detection. The possibility of speci cation based con ict resolution through withdrawal allows users to state inheritance preference. We present a formal account of the semantics of our language by de ning a model theory, proof theory and a xpoint theory. We also prove that the di erent characterizations of our language are equivalent. iii Abstract We nally present an elegant technique to reduce inheritance to deduction based on the idea of constrained deduction, called the i-completion. The reduction technique makes it possible to implement object-oriented databases with inheritance, overriding and con ict resolution in a purely deductive system. An ORLog prototype implementation on Coral deductive database system is discussed based on this reduction technique. We are able to exploit the rich set of query optimization techniques available in Coral since the implementation does not require meta-interpretation. iv Dedication
To My Parents and To My Grandfather Abdul Hai Acknowledgments
I have been fortunate to have Dr. Laks V. S. Lakshmanan and Dr. Fereidoon Sadri as my thesis supervisors. They always encouraged me to pursue the best and strive for excellence. I was introduced to the challenging eld of logic and object-orientation by Dr. Lakshmanan and he gave me the foundation I needed to advance in my research. Despite a busy schedule, he always had the time for me, and his guidance and our numerous discussions paved the way for the completion of this thesis. I thank him for taking the time to read the manuscript of this thesis and xing several bugs. I would like to express my indebtedness to him for his exceptional guidance and inspiration. Fereidoon gave me the basics of databases and rudimentary details of deductive databases at the early stage of my research. The interest he developed in me later triggered my fascination in deductive object-oriented databases. He then introduced me to uncertainty management in databases, and always encouraged me to explore new frontiers in research. It was a real delight doing research with him in these areas. He always had faith in me, took interest in me and helped me in shaping my professional life. I am grateful to him for everything he did for me. I was also blessed by two great souls, Dr. V. S. Alagar and Dr. Peter Grogono, during my stay at Concordia. They always gave me the much needed encouragement, and helped me change my views towards teaching, research and life as a whole. I received so...
View Full Document