Applying Case-Based Reasoning: Techniques for Enterprise Systems (The Morgan Kaufmann Series in Artificial Intelligence) - Softcover

Watson, Ian

Softcover

ISBN 10: 1558604626 ISBN 13: 9781558604629

Publisher: Morgan Kaufmann, 1997

Synopsis

Case-based reasoning (CBR) is an intelligent-systems method that enables information managers to increase efficiency and reduce cost by substantially automating processes such as diagnosis, scheduling and design. A case-based reasoner works by matching new problems to "cases" from a historical database and then adapting successful solutions from the past to current situations. Organizations as diverse as IBM, VISA International, Volkswagen, British Airways, and NASA have already made use of CBR in applications such as customer support, quality assurance, aircraft maintenance, process planning, and decision support, and many more applications are easily imaginable.

It is relatively simple to add CBR components to existing information systems, as this book demonstrates. The author explains the principles of CBR by describing its origins and contrasting it with familiar information disciplines such as traditional data processing, logic programming, rule-based expert systems, and object-oriented programming. Through case studies and step-by-step examples, he goes on to show how to design and implement a reliable, robust CBR system in a real-world environment. Additional resources are provided in a survey of commercially available CBR tools, a comprehensive bibliography, and a listing of companies providing CBR software and services.

About this title

Review

Case-based reasoning (CBR) is one of the most promising recent technologies for building intelligence into computers. This useful introduction to the technology shows what CBR is, why it is important, and how it can be applied to some intriguing real-world problems. The early part of this text outlines the the so-called four "re-'s" of CBR: retrieve, reuse, revise, and retain. Basically, CBR systems match possible solutions to a database of earlier questions, retrieve the answer, and adapt (or revise) it. If appropriate, CBR systems can even save (or retain) these new problems and solutions to its "case-base" for later reuse. For example, if your car does not start, you can use a CBR to match the characteristics of your problem with a database of diagnoses.

The author does a fine job of distinguishing CBR from other artificial intelligence models, such as expert systems (which are rules-based and can be hard to develop because experts don't always get their expertise translated perfectly by programmers) and neural networks (CBR can be more accurate because it can match characteristics and give a reason, while a neural network cannot), and proving that CBR can be faster computationally. He also gives an excellent presentation on how CBR has been used successfully in a surprising range of applications, such as manufacturing, medicine, law, and help desk systems for products, including Web-based applications systems for users. (Help desk applications are a natural fit for CBR technology, but they are not its only use.)

CBR has already enjoyed considerable success without many people knowing about it. (As the author suggests, expert systems, a promising older artificial intelligence [AI] technology, never quite lived up to its initial dazzle.) The author shows that CBR research has already been used to create products. He provides a survey of tools from two continents as well as an extensive bibliography for those experts who want to read more on the subject.

About the Author

Ian Watson is Senior Lecturer in Intelligent Systems at the University of Salford in the United Kingdom, and one of the world's most active researchers in CBR. He became interested in CBR after realizing its promise for industrial applications and its comparative ease of implementation. He is a member of the Research Centre for the Built and Human Environment (BUHU), where he continues his investigations of intelligent-systems methodologies and object-oriented programming and maintains AI-CBR, the center for CBR information on the World Wide Web.

Edition details

• Publisher: Morgan Kaufmann
• Publication date: 1997
• ISBN 10: 1558604626
• ISBN 13: 9781558604629
• Binding: Paperback
• Language: English
• Edition number: 1
• Number of pages: 290