Building Expert Systems in Training and Education

Grabinger, Wilson, Jonassen

Praeger Publishers

New York

1990

REVIEWED BY: Danny McCoy

DATE: 8/98

Building Expert Systems in Training and Education was written to give instructional systems designers a look into incorporating experts systems into instructional design. The book attempts to demonstrate the process of developing small- scale expert systems that instructional designers can run from desktop computers.

The book attempts to remain general so as not to be hardware or software specific.

The author follows an instructional systems design model to present his material. The book begins with an excellent chapter on what an expert system is. He defines an expert system as, " practical tools that can server as intelligent job aids to facilitate on-the-job decision making in tasks such as judging projects, diagnosing learning problems, identifying and classifying performance problems, or helping consumers to decide among a large number of alternatives. Expert systems are tools that can be used to improve human performance." (p.1) After an introduction to expert systems the author provides seven tasks for the expert system design process. They are Identification of an appropriate problem, limitation of the problem domain, specification of problem solutions/goals, specification of problem attributes and attribute values, solution matrix, generalization of rules and examples, and implementation. Each is covered in detail in a separate chapter. This format follows the typical instructional systems design model by running a needs assessment, defining goals, writing instruction and application of the material. The author tries to provide you with information on how to assess the need for an expert system. He then explains that you start from what you want the expert system to accomplish and work backward to design rules. He explains how heuristic expert systems operate on if/then rules with an inference engine to come to a conclusion. His purpose is to provide the reader with the basic idea of how expert systems work so that the instructional designer can use a shell software to create their own expert system that can be run on desktop computers. Shell software provides an interface that allows the designer to write rules but does not involve the complicated codes.

The author takes a weak Artificial Intelligence point of view. By weak, it is meant that computer only act like humans they do not truly think (strong) like humans. In the author attempt to explain artificial intelligence he breaks up the words and comments on artificial as defined as simulated. He contends, "the word simulate is a critical part of this description of AI because it helps us to narrow our expectations for both artificial intelligence and expert systems. A simulation is something that looks real and feels real but is not real." Then entire book looks at how to ability of the computer to be intelligent is dependant on the rules created by the designer. Never once mentioned, as is the case with many in the field of AI, is the computer being able to think on it’s own and learn like a human. The author makes it clear that these systems can be build to appear to do what we as humans would consider intelligent in humans. He stress that expert systems must be narrow in domain to be able to appear intelligent.

The book can be helpful to the instructional systems designer in that it shows another tool to help design instruction. With a basic knowledge on how expert systems work, software can be purchased at reasonable prices that will help the designer write his own expert system. One major flaw with the book is that is does not focus on education as the title would suggest. Only in the last few pages though does he give any application of the use of expert systems in education. These examples are very short. The main example of an expert system use throughout the book is one built to help a car salesman select the best car for his customer based on a large range of options. This example was an excellent one to show how to build an expert system but it misses what one would believe the point of the book to be, Building Expert Systems in Training and Education.