• Expert systems are knowledge-based systems and solve problems in a broader perspective.
It is an application of artificial intelligence that aims to imitate human intelligence in examining and solving
is the area. In this imitation, algorithms and inference mechanisms interact and
It works.
• To give a more specific definition, based on information received from an expert
It is a computer aided decision tool.
• These intelligent systems solve problems in a field by thinking like an expert.
and by learning the problem it solves, storing it in the knowledge store and using what you have learned before
without making a structural change in the information that results in a change in the structure of the expert system.
It ensures that there is no need for change.
• In creating the expert system, the knowledge engineer receives the necessary information from the domain expert.
It needs to retrieve and translate them in a way that is stored in the knowledge base.
• Expert systems use experience-based inference methods instead of precise and clear algorithms.
For this reason, the design of expert systems is complex and time-consuming.
is the process.
• Knowledge Base; It is a database where experience-based information specific to the relevant field is stored. rule and
It consists of facts. Facts; relationships between objects, limitations and explanations
occurs. The rules are; defines logical relationships between concepts related to the problem area.
• Reasoning Unit; By reading the rules and facts, you understand what they mean and
It performs the reasoning function.
• User interface; It provides communication between the user and the system. Usually, Why and
It includes an explanation unit that answers how questions.
• Information Retrieval Unit; Allows the user to correct, add and edit rules and facts in the knowledge base.
Allows you to make stickers and delete some of them.
• Description Unit; Explains how reasoning is done. In addition, communication with the user
It asks some questions and the user wants to know why he is asking this question. Explanation
The unit makes the necessary explanation.
One of the general techniques of expert systems is the character and word matching technique.
Any change in the knowledge base of a developed system does not affect the entire system.
There are decision units to make decisions on their own. These and similar features, Expert
It makes the systems different from other programs. For this reason, software usually uses PROLOG
and LISP programming languages.
GENERAL STRUCTURE OF EXPERT SYSTEM PROGRAMS
• Expert system programs generally include Reasoning; That is, according to the available data, the most
They work on the basis of determining the appropriate situation. Usually all rules in the Knowledge Base
Reasoning is done using two techniques.
• Forward Chaining
• Chaining Backward
Forward Chaining
• The reasoning unit starts from the beginning of the problem (from the IF sentence) to the conclusion.
(THEN...) is to reach. This method works with the logic of induction. Condition of all rules
The conclusion is reached by taking into account whether it provides or not. If the conditions are met, "Then"
The judgment sentence in the section is correct. This sentence is the result obtained according to the conditions.
Chaining Backward
• Reasoning unit; While solving the problem, use the result (THEN...) sentence, which is the end of the rule.
begins and a solution is found by applying conditional (IF...) sentences. So this kind of chaining
It takes the principle of deduction as a basis and uses all the rules that will provide the conclusion part one by one.
examines.
HUMAN FACTOR IN EXPERT SYSTEMS
• At least two people are involved in the development and use of Expert Systems; expert and
user. Often the knowledge engineer and system builder are also included in this group.
• Expert: Along with his special knowledge, judgment, experience and method, the expert uses these abilities to solve problems.
can implement and give advice. The task of the expert is to determine the tasks that the information system will perform.
is to present to the system how to do it. The expert determines which facts are important and
He is the person who knows the meaning of the relationships between facts.
• User: USs have many users. They may appear with the following identities:
A non-expert client who wants direct advice, a student who wants to learn, a
US makers, experts, etc. who want to develop or increase their base.
• Users may not have knowledge about computers or in-depth knowledge of problems.
But many people use USs to get a quicker and probably less costly service.
They want to reach decisions. The capabilities of USs were developed to save time and effort.
For, these provide users with the shortest answers unlike traditional computer systems.
• Knowledge Engineer: The knowledge engineer helps the expert to structure problem areas.
You can do this by interpreting human-expert answers, integrating them, making analogies, reversing them.
He does this by giving examples. This person is also (mostly) the person who makes the system. in US construction
The lack (scarcity) of knowledge engineers is a significant problem. Those who designed the US faced this difficulty
They reduce the need for knowledge engineers by using production tools to defeat
• Other Participants: Many other people can participate in USs. For example, the system constructor uses other USs
It helps integrate with computer systems. The seller and other support staff also
They can be expressed under the title.
TODAY'S LIMITATIONS OF EXPERT SYSTEMS
• Expert Systems have insufficient connections with the outside world.
• Grassroots knowledge is superficial.
• It is extremely dependent on human-expert labor.
• Almost no learning skills.
• Usage area may be considered limited for now.
• Methods of reasoning are limited.
• Information presentation methods are limited.
ADVANTAGES AND DISADVANTAGES OF EXPERT SYSTEMS
• Advantages;
• Since expert systems are faster than human experts, this speed difference increases production.
may be reflected. This speed also allows the production cycle to be completed in a shorter time and
It is important to maintain the continuity of the system while doing
can be seen. To maintain continuity, there is no loss of data and information over time.
• If an expert system can be put in place instead of experts, the cost incurred for experts
disappears and therefore indirectly production increases. Experts instead of expert people
Another benefit of using these systems is the increase in the quality level.
• For the purpose of designing expert systems, all the characteristics of a human
It also includes expressions that do not indicate certainty, such as "I'm not sure" and "I don't know".
giving, evaluating and adjusting the criteria accordingly, that is, with complete and imprecise information
workability can be achieved.
• Expert systems are some of the systems in which these systems are used in order to display intermediate results.
The aim is to see interim results in the fields in reports and to make plans and designs for the future.
It can be designed so that intervention can be made at any time.
• Expert systems will contribute to the development of the factories of the future and accordingly
The planned unmanned factories will be further developed with expert systems. Especially
A major breakthrough is expected in this regard.
• Disadvantages;
• In today's conditions, there cannot be enough experts in every field, and even if they are, there is not enough time.
This reveals the lack of expert knowledge, that is, the difficulty of accessing this knowledge.
puts it. In addition, difficulties in obtaining information from experts are another problem.
arises because some experts cannot convey what they know.
• As a result of experts bringing different perspectives to the same subject, it depends on the field to be applied.
Although it varies, the information adopted and presented correctly by experts
It may be inconsistent or it may have a very economical yet practical result.
it could be. Overcoming such problems can only be achieved by creating a dynamic expert system design and
Depending on the situation, it may be possible to reveal real knowledge as a result of acquiring the ability to learn.
• High cost and excess time manifest as problems, the reason for this is expert
the high cost of designing the system and, in addition, the need for knowledge engineers, experts and programmers.
These are the costs of the systems that vary between 3-5 years.
• Since creativity is a uniquely human skill, it can be used when faced with an unusual situation.
While a new solution can be offered with the creativity of an expert, the expert system cannot do this.
However, although learning skills can be provided to expert systems, this system
Designing poses a problem in itself, but an expert can learn
This skill is realized very easily.
USAGE AREAS OF EXPERT SYSTEMS
• Interpretation: Identification of states from sensor transmitters
• Area of Use: Voice recognition, Image Analysis, Control
• Prediction: Drawing conclusions similar to given situations.
• Area of Use: Weather forecast, Grain Forecast
• Diagnosis: Detection of system disorders according to observation results
• Area of Use: Medicine, Electronics
• Design: Object design under limited conditions
• Area of Use: Circuit Drawing
• Planning: Design of operations
• Area of Use: Automatic Programming, Military Planning
• Visualization: Comparing observations to plan precisions
• Area of Use: Regulation and Cost Management of Nuclear Power Plants
• Debugging: Presenting defects that cause errors
• Area of Use: Computer Software
• Repair: Execution of the specified management plan
• Area of Use: Automobile, Computer
• Education: Detection and correction of student behavior
• Area of Use: Counseling, Rehabilitation, Treatment
• Control: Interpretation, prediction, repair and monitoring of System Behavior
• Area of Use: Air Traffic Control, Combat Control
