The inference engine

In rule-based systems, an interpreter module controls the application of the rules and, thus, the systems activity. In a basic cycle of activity (i.e., recognize-act cycle) 

There are two prerequisites for the application of pattern methods. First, the patterns are retrieved from calculated data, and, thus, the accuracy and reliability of a pattern in a given context must be validated. In terms of descriptors, the accuracy relies mainly on the experimental conditions or the raw data used for calculation. 

The second requirement is a suitable similarity measure for the comparison of patterns. Once the patterns are defined and the quality of the experimental base data is good, pattern-recognition methods are valuable. Nevertheless, if patterns change irregularly and cannot be explicitly defined, the similarity measure no longer describes the difference between query and experimental pattern even if a fuzzy logic approach is implemented. 

In general, systems based on the comparison of patterns can provide a series of candidates, and, finally, the expert has to decide the target compound by experience or by using additional information. 

Rules are defined by subject-matter experts using a high-level rules language. They are collected into rule sets that are then translated at run time into an executable Rete. 

---

Inference engine The inference engine represents the central problem-solving subsystem. It contains strategies for using the information contained in the... 

The typical structure of an expert system is shown in Fig. 43.1. Three basic components are present in all expert systems the knowledge base, the inference engine and the interaction module (user interface). The knowledge base is the heart of the expert system. It contains the necessary expert knowledge and experience to act as a decision support. Only if this is correct and complete enough the expert system can produce meaningful and useful conclusions and advice. The inference... 

Rules seemingly have the same format as IF.. THEN.. statements in any other conventional computer language. The major difference is that the latter statements are constructed to be executed sequentially and always in the same order, whereas expert system rules are meant as little independent pieces of knowledge. It is the task of the inference engine to recognize the applicable rules. This may be different in different situations. There is no preset order in which the rules must be executed. Clarity of the rule base is an essential characteristic because it must be possible to control and follow the system on reasoning errors. The structuring of rules into rule sets favours comprehensibility and allows a more efficient consultation of the system. Because of the natural resemblance to real expertise, rule-based expert systems are the most popular. Many of the earlier developed systems are pure rule-based systems. 

In this example this combination is straightforward. One has to be aware, however, that a medium-sized expert system easily contains several hundreds of rules. In addition, several rules can be valid at the same time. The inference engine should also have a strategy for deciding on priorities conflict resolution). 

An interpreter, or explanation system, forms the interface between the user and the core software. It interprets queries from the user so that the inference engine can make sense of them and presents the conclusions of the inference engine in language appropriate to the needs and level of knowledge of the user. It is also responsible for generating explanations of the reasoning that... 

In backward chaining, the inference engine would search through the knowledge base until it found a rule, the conclusion of which was "Go to live with mother." It would find three such rules ... 

In a goal-driven system, the inference engine would start with the first rule because the conclusion of that rule is the recommendation of a solvent. 

The user interface, the interpreter, and the inference engine together comprise the ES shell or a skeletal system. As these components are largely independent of the specific application, all that is needed to create a working ES from the shell is to feed it with rules and facts. 

Within a fuzzy system, an inference engine works with fuzzy rules it takes input, part of which may be fuzzy, and generates output, some or all of which may be fuzzy. Although the role of a fuzzy system is to deal with uncertain data, the input is itself not necessarily fuzzy. For example, the data fed into the system might consist of the pH of a solution or the molecular weight of a compound, both of which can be specified with minimal uncertainty. In addition, the output that the system is required to produce is of more value if it is provided in a form that is crisp "Set the thermostat to 78°C" is more helpful to a scientist than "raise the temperature of the oven." Consequently, the fuzzy core of the inference engine is bracketed by one step that can turn crisp data into fuzzy data, and another that does the reverse. 

When the system is used to diagnose the power plant chemistry, the inference engine will activate all the rules for which evidence exists. Thus all possible conclusions are examined... 

If the method for acquiring a VALUE is ASKIT, then a user PROMPT is stored. In order to guarantee a valid response to the question, a LISP function to check the answer is included with the FACT. Table I lists the currently implemented response checking functions. Whenever the inference engine reaches one of these facts, searching is stopped and the user is prompted for a value. 

This TEXT is used in the various explanation and tracing facilities. Whenever the inference engine reaches one of these FACTS it either continues its search, if possible, or proceeds another level deeper in the reverse search and tries to prove that FACT. 

The inference engine was designed to use multivalued logic, i.e., it handles inexact reasoning. Confidence factors (CF) are contained in the THEN clauses of each rule. The equation for combining positive confidences is ... 

Most of the applications of artificial intelligence in chemistry so far have not involved numerical computation as a primary goal. Yet there are aspects of the AI approach to problem-solving which have relevance to computation. In scientific computation, one could view the knowledge base as the set of equations, input variable values, and unit conversions relevant to the problem, and the inference engine the numerical method used to solve the equations. This paper describes such a software system,... 

SpinPro is a typical backward chaining, rule-based expert system. Rule-based systems are systems in which the expert s knowledge is encoded primarily in the form of if-then rules, i.e., if a set of conditions are found to be true then draw a conclusion or perform an action. "Backward chaining" refers to the procedure for finding a solution to a problem. In a backward chaining system, the inference engine works backwards from a hypothesized solution to find facts that support the hypothesis. Alternative hypotheses are tried until one is found that is supported by the facts. 

---