Expert Systems For Software Engineers And Managers

This book is written for software engineers, software project leaders, and software managers who would like to introduce a new advanced software technology, expert systems, into their product. Expert system technology brings into programming a new dimension in which "rule of thumb" or heuristic expert knowledge is encoded in the program. In contrast to conventional procedural languages {e. g. , Fortran or C}, expert systems employ high-level programming languages {Le. , expert system shells} that enable us to capture the judgmental knowledge of experts such as geologists, doctors, lawyers, bankers, or insurance underwriters. Past expert systems have been more successfully applied in the problem areas of analysis and synthesis where the boundary of lo;nowledge is well defined and where experts are available and can be identified. Early successful applications include diagnosis systems such as MYCIN, geological systems such as PROSPECTOR, or design/configu ration systems such as XC ON. These early expert systems were mainly applicable to scientific and engineering problems, which are not theoreti cally well understood in terms of decisionmaking processes by their experts and which therefore require judgmental assessment. The more recent expert systems are being applied to sophisticated synthesis problems that involve a large number of choices, such as how the elements are to be compared. These problems normally entailed a large search space and slower speed for the expert systems designed. Examples of these systems include factory scheduling applications such as ISIS, or legal reasoning applications such as TAXMAN.

This report was prepared to assist managers in the risk-managed application of expert systems (ES) technology. Two problems that have frustrated managers and technicians who might otherwise be interested in exploring ES solutions are the seeming lack of a software engineering discipline for creating and maintaining expert systems and the fact that the software development risks associated with expert systems have not been clearly addressed. This guide focuses on ES development as a software engineering endeavor and emphasizes that, as such, it should be subject to much the same discipline as conventional automated information system software. The guide is meant to be used in conjunction with accepted software development methods and standards, such as Department of Defense (DOD) Standards 2167A and 7935.1, DOD Directive 7920.1, DOD Instruction 7920.2, and the OSD Major Automated System Review Council (MAISRC) Guidelines for Program Managers. This document presents a step-by-step guideline that addresses the managerial and technical risks of each phase in the software development life cycle of systems that contain ES components.

The aim of this volume is to give practical guidance on how to establish and use expert systems technology within an organization, avoiding various pitfalls on the way. The book also outlines the resulting benefits in terms of productivity.

Managers, business owners, computer literate individuals, software developers, students, and researchers--all are looking for an understanding of artificial intelligence (AI) and what might be in the future. In this literate yet easy-to-read discussion, Derek Partridge explains what artificial intelligence can and cannot do, and what it holds for applications such as banking, financial services, and expert systems of all kinds. Topics include: the strengths and weaknesses of software development and engineering; machine learning and its promises and problems; expert systems and success stories; and practical software through artificial intelligence.

This book presents current research in the study of expert system software, including algorithms based on paraconsistent annotated logic in expert systems; expert flow systems in the agronomical, industrial, environmental, pharmaceutical and geological fields; expert systems in remote sensing; domain-wide expert system applications; modeling a parser as an expert system; embedded expert systems for flow control of delay sensitive real-time traffic in WLANs; expert systems in fund-raising management and algebraic approaches in the development of rule based expert systems.

Managing Expert Systems explores the trends in expert systems development and implementation. As top authorities in the field of ES, Turban and Liebowitz examine the factors that contribute to the development of a successful expert system.

The first book to discuss efficient ways to implement the systems currently being developed--written by the co-author of Expert Systems: Artificial Intelligence in Business, generally regarded as the best non-technical guide to expert systems for business people. Gives innovative ideas for using expert systems to facilitate business operations. Appropriate as a text or supplement for data base, decision support, or special-topic courses that cover expert systems. Clearly explains new applications of automatic decision-making in management, sales, operations, programming, research, and service industries. Text supported by extensive examples and graphs.