AN EQUATIONAL APPROACH TO DECISION SUPPORT SYSTEMS (KNOWLEDGE, PROBLEM-PROCESSING, LANGUAGE, TEMPORAL SEMANTICS)
Abstract
Decision support system designers have long been relying on different software to design decision support systems. However, those software often presents different expertises. This thesis investigates the use of equations as an independent tool to formalize decision support systems. First, a many-sorted equational formalism is built. Such a formalism preserves the Soundness and Completeness. Examples are constructed to demonstrate the generality of equational formalism and the replacement mechanism. Second, an equational database model is suggested to accommodate relational, hierarchical and network models. Such an equational data model is built upon primitive information represented as a set of equations. As such, diverse users' views can always be derived. Finally, by recognizing time as a fundamental concept to model a real environment, a point and interval temporal model is proposed to help construct a knowledge space for a decision support system. In order to make systems directly available to decision makers, a natural language grammar is developed and included into the knowledge space. As such, natural language processing and temporal reasoning can be treated in an integrated manner. The issue of processing natural language queries with explicit temporal references is also addressed.
Degree
Ph.D.
Subject Area
Computer science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.