The database semantics of time

A database management system is a tool for manipulating large databases consisting of special software for the input, interrogation, maintenance, and analysis of data. The conceptual level in a database management system is a specification of the semantics of the database in terms of abstract entities and relationships existing in the real world system being described. The conceptual level model developed in this dissertation is an extension of the binary-based model of J. R. Abrial that accommodates a generalization hierarchy; this model is called the Extended Abrial Conceptual Level (EACL) model. Time is a universal attribute of the real world and, hence, of information about the world. Since real world systems vary in time, a realistic conceptual level model should capture this dynamic quality. An explicit description of discrete time, called the Discrete Time Domain Generalization, is given using the EACL language. The Discrete Time Domain Generalization is intended to be an integral part of a database management system designed to manage time-dependent information. It is a high level, semantic description of the structure and behavior of a wide variety of time systems. It is also an extensible model of time since the mechanism for defining new time domains is also part of the conceptual level. The Discrete Time Domain Generalization is used to examine and define consistency constraints for time-dependent information in a database. For example, by introducing a graphical description of a process at the conceptual level, it is possible to constrain the chronological order of event sequences for the process. Other consistency constraints such as event duration and event mutual exclusion are also presented.