DEFINITION - An expert system is a computer program that simulates the judgement and behavior of a human or an organization that has expert knowledge and experience in a particular field. Typically, such a system contains a knowledge base containing accumulated experience and a set of rules for applying the knowledge base to each particular situation that is described to the program. Sophisticated expert systems can be enhanced with additions to the knowledge base or to the set of rules.
Among the best-known expert systems have been those that play chess and that assist in medical diagnosis.
An expert system is software that attempts to provide an answer to a problem, or clarify uncertainties where normally one or more human experts would need to be consulted. Expert systems are most common in a specific problem domain, and is a traditional application and/or subfield of artificial intelligence (AI). A wide variety of methods can be used to simulate the performance of the expert; however, common to most or all are: 1) the creation of a knowledge base which uses some knowledge representation structure to capture the knowledge of the Subject Matter Expert (SME); 2) a process of gathering that knowledge from the SME and codifying it according to the structure, which is called knowledge engineering; and 3) once the system is developed, it is placed in the same real world problem solving situation as the human SME, typically as an aid to human workers or as a supplement to some information system. Expert systems may or may not have learning components.
The MYCIN rule-based expert system introduced a quasi-probabilistic approach called certainty factors, whose rationale is explained below.
A human, when reasoning, does not always make statements with 100% confidence: he might venture, "If Fritz is green, then he is probably a frog" (after all, he might be a chameleon). This type of reasoning can be imitated using numeric values called confidences. For example, if it is known that Fritz is green, it might be concluded with 0.85 confidence that he is a frog; or, if it is known that he is a frog, it might be concluded with 0.95 confidence that he hops. These certainty factor (CF) numbers quantify uncertainty in the degree to which the available evidence supports a hypothesis. They represent a degree of confirmation, and are not probabilities in a Bayesian sense. The CF calculus, developed by Shortliffe & Buchanan, increases or decreases the CF associated with a hypothesis as each new piece of evidence becomes available. It can be mapped to a probability update, although degrees of confirmation are not expected to obey the laws of probability. It is important to note, for example, that evidence for hypothesis H may have nothing to contribute to the degree to which Not_h is confirmed or disconfirmed (e.g., although a fever lends some support to a diagnosis of infection, fever does not disconfirm alternative hypotheses) and that the sum of CFs of many competing hypotheses may be greater than one (i.e., many hypotheses may be well confirmed based on available evidence).
The CF approach to a rule-based expert system design does not have a widespread following, in part because of the difficulty of meaningfully assigning CFs a priori. (The above example of green creatures being likely to be frogs is excessively naive.) Alternative approaches to quasi-probabilistic reasoning in expert systems involve fuzzy logic, which has a firmer mathematical foundation. Also, rule-engine shells such as Drools and Jess do not support probability manipulation: they use an alternative mechanism called salience, which is used to prioritize the order of evaluation of activated rules.
In certain areas, as in the tax-advice scenarios discussed below, probabilistic approaches are not acceptable. For instance, a 95% probability of being correct means a 5% probability of being wrong. The rules that are defined in such systems have no exceptions: they are only a means of achieving software flexibility when external circumstances change frequently. Because rules are stored as data, the core software does not need to be rebuilt each time changes to federal and state tax codes are announced.
Two methods of reasoning when using inference rules are forward chaining and backward chaining.
Forward chaining starts with the data available and uses the inference rules to extract more data until a desired goal is reached. An inference engine using forward chaining searches the inference rules until it finds one in which the if clause is known to be true. It then concludes the then clause and adds this information to its data. It continues to do this until a goal is reached. Because the data available determines which inference rules are used, this method is also classified as data driven.
Backward chaining starts with a list of goals and works backwards to see if there is data which will allow it to conclude any of these goals. An inference engine using backward chaining would search the inference rules until it finds one which has a then clause that matches a desired goal. If the if clause of that inference rule is not known to be true, then it is added to the list of goals.
The following general points about expert systems and their architecture have been outlined:
1. The sequence of steps taken to reach a conclusion is dynamically synthesized with each new case. The sequence is not explicitly programmed at the time that the system is built.
2. Expert systems can process multiple values for any problem parameter. This permits more than one line of reasoning to be pursued and the results of incomplete (not fully determined) reasoning to be presented.
3. Problem solving is accomplished by applying specific knowledge rather than specific technique. This is a key idea in expert systems technology. It reflects the belief that human experts do not process their knowledge differently from others, but they do possess different knowledge. With this philosophy, when one finds that their expert system does not produce the desired results, work begins to expand the knowledge base, not to re-program the procedures.
There are two styles of user-interface design followed by expert systems. In the original style of user interaction, the software takes the end-user through an interactive dialog. In the following example, a backward-chaining system seeks to determine a set of restaurants to recommend:
Q. Do you know which restaurant you want to go to?
Q. Is there any kind of food you would particularly like?
Q. Do you like spicy food?
Q. Do you usually drink wine with meals?
Q. When you drink wine, is it French wine?
There are generally three individuals having an interaction in an expert system. Primary among these is the end-user, the individual who uses the system for its problem solving assistance. In the construction and maintenance of the system there are two other roles: the problem domain expert who builds the system and supplies the knowledge base, and a knowledge engineer who assists the experts in determining the representation of their knowledge, enters this knowledge into an explanation module and who defines the inference technique required to solve the problem. Usually the knowledge engineer will represent the problem solving activity in the form of rules. When these rules are created from domain expertise, the knowledge base stores the rules of the expert system.
An understanding of the "inference rule" concept is important to understand expert systems. An inference rule is a conditional statement with two parts: an if clause and a then clause. This rule is what gives expert systems the ability to find solutions to diagnostic and prescriptive problems. An example of an inference rule is:
If the restaurant choice includes French and the occasion is romantic,
Then the restaurant choice is definitely Paul Bocuse.
Procedure node interface
The function of the procedure node interface is to receive information from the procedures coordinator and create the appropriate procedure call. The ability to call a procedure and receive information from that procedure can be viewed as simply a generalization of input from the external world. In some earlier expert systems external information could only be obtained in a predetermined manner, which only allowed certain information to be acquired. Through the knowledge base, this expert system disclosed in the cross-referenced application can invoke any procedure allowed on its host system. This makes the expert system useful in a much wider class of knowledge domains than if it had no external access or only limited external access.
In the area of machine diagnostics using expert systems, particularly self-diagnostic applications, it is not possible to conclude the current state of "health" of a machine without some information. The best source of information is the machine itself, for it contains much detailed information that could not reasonably be provided by the operator.
The knowledge that is represented in the system appears in the rulebase. In the rulebase described in the cross-referenced applications, there are basically four different types of objects, with the associated information:
1. Classes: Questions asked to the user.
2. Parameters: Place holders for character strings which may be variables that can be inserted into a class question at the point in the question where the parameter is positioned.
3. Procedures: Definitions of calls to external procedures.
4. Rule nodes: Inferences in the system are made by a tree structure which indicates the rules or logic mimicking human reasoning. The nodes of these trees are called rule nodes. There are several different types of rule nodes.