1 – introduction
In this chapter, we tend to introduce the concepts behind production systems, or skilled systems, and make a case for however they will be engineered exploitation rule-based systems, frames, or a mix of the 2.
This chapter explains techniques like forward and backward chaining, conflict resolution, and also the anatomical structure algorithmic program. It conjointly explains the design of AN skilled system and describes the roles of the people WHO square measure concerned in planning, building, and exploitation skilled systems.
2 – Rules for data illustration
One way to represent data is by exploitation rules that specific what should happen or what will happen once bound conditions square measure met. Rules square measure sometimes expressed within the sort of IF . . . THEN . . . statements, such as:
IF A THEN B
This can be thought of to own an identical logical that means because the following:
A → B
As we tend to saw in Chapter seven, A is named the antecedent and B is that the resulting during this statement. In expressing rules, the ensuing sometimes takes the shape of AN action or a conclusion. In alternative words, the aim of a rule is sometimes to inform a system (such as AN skilled system) what to try to to in bound bound, or what conclusions to draw from a group of inputs regarding the present scenario.
In general, a rule will have quite one antecedent, sometimes combined either by AND or by OR (logically a similar because the because the because the we tend to saw in Chapter 7). Similarly, a rule could have quite one resulting, that sometimes suggests that there square measure multiple actions to be taken.
In general, the ANtecedent of a rule compares an object with a potential worth, exploitation AN operator. as an example, appropriate antecedents in an exceedingly rule may be
IF x > 3
IF name is “Bob”
IF weather is cold
Here, the objects being thought of square measure x, name, and weather; the operators square measure “>” and “is”, and also the values square measure three, “Bob,” and cold. Note that AN object isn’t essentially AN object within the real-world sense—the weather isn’t isn’t object, however rather a state or condition of the globe. AN object during this sense is solely a variable that represents some entity or state within the planet.
An example of a rule may be
IF name is “Bob” AND weather is cold THEN tell Bob ‘Wear a coat’
This is AN example of a recommendation rule, that takes a group of inputs and provides recommendation as a result. The conclusion of the rule is truly AN action, and also the action takes the shape of a recommendation to Bob that he ought to wear a coat. In some cases, the foundations give additional definite actions like “move left” or “close door,” during which case the foundations square measure being employed being employed directives.
Rules may be wont to represent relations such as:
IF temperature is below zero
THEN weather is cold
3 – Rule-Based Systems
Rule-based systems or production systems square measure pc systems that use rules to produce recommendations or diagnoses, or to work out a course of action in an exceedingly specific scenario or to unravel a selected drawback.
A rule-based system consists of variety of components:
- a info of rules (also referred to as a data base)
- a info of facts
- AN interpreter, or reasoning engine
In a rule-based system, the cognitive content consists of a group of rules that represent the data that the system has. The info of facts represents inputs to the system that square measure wont to derive conclusions, or to cause actions.
The interpreter, or reasoning engine, is that the a part of the system that controls the method of explanation conclusions. It uses the foundations and facts, and combines them along to draw conclusions.
As we’ll see, these conclusions square measure usually derived exploitation deduction,although there square measure alternative potential approaches. exploitation deduction to achieve a conclusion from a group of antecedents is named forward chaining. AN AN technique, backward chaining, starts from a conclusion and tries to point out it by following a logical path backward from the conclusion to a group of antecedents that square measure within the info of facts.