Expert Systems: Key Components and Applications
Expert Systems: These are systems that solve problems using a symbolic representation of human knowledge. They contain a logical unit with which conclusions are drawn from the knowledge base. The system decides which rules to apply, what should be done, resolves conflicts, and makes use of the knowledge base.
Inference Mechanism Functions
The inference mechanism determines the actions that take place, the order in which they will occur, and how they will affect the different parts of the Expert System. It also determines how and when rules are processed and controls the dialogue with the user.
Search Types Implemented
Factors influencing search type include how the language has been written, working speed, and strategic solutions (unordered: random, heuristic; Computer: forward and backward chaining). Other factors include the way you choose to represent knowledge, whether to include metacognition, the kind of logic used in reasoning, and the method for evaluating incomplete or uncertain knowledge.
Knowledge Base
The Knowledge Base represents the universe where the system is inserted. It contains all the facts, rules, and procedures of the application domain that are important to solve the problem.
Characteristics of the Interface
- Learning management should be fast.
- The user should not spend much time managing the system.
- The system should be intuitive and easy to handle.
- It should not be forgotten that the Expert System simulates the behavior of an expert.
- The system must be comfortable and relatively simple to handle.
- Erroneous data entry should be avoided as far as possible.
- Results should be presented clearly.
- User questions and explanations should be understandable.
Acquisition Subsystem
This subsystem may focus primarily on the structuring of knowledge without having to spend so much time on programming activity. Requirements include clear knowledge representation possibilities, automatic syntax checking, and constant possibility of access to the programming language.
Subsystem Explanations
It is always desirable that the system development work is known, as well as the extent of progress in processing the problem.
The Basis of Facts
Facts differ from data in the sense that the events are part of the System-Based Controller (SBC), while the data can vary from one solution to another and should be grouped into external files to the SBC.
The Shell
The Shell is an expert system that contains an empty knowledge base, separating the knowledge base from the procedures that use the knowledge.
Types of Expert Systems and Their Applications
Prediction System
Infers the likely consequences from a given set of circumstances.
- Population
- Traffic
- Damage to crops by some kind of insect, pest, etc.
- Armed conflicts based on intelligence reports
- Demand for some input (oil, copper), given a situation (geopolitical, economic, development)
Diagnostic Systems
Associates irregularities with observed behavior and possible causes.
- Disease from a set of symptoms
- Defective components of a system
- Equipment failures in production processes
Planning System
These systems are intended to design action plans.
- Programming routes
- Programming robots
- Communications programming
- Programming experiments
- Creating flight plans
Monitoring Systems
Compares the behavior of a system with expected behavior. From the differences, it suggests corrective actions.
- Monitor readings of instruments to detect fault conditions of industrial equipment
- Detect conditions conducive to accidents
- Monitoring of air traffic
Debugging Systems
Systems to find appropriate remedies for malfunctions.
- Select type of maintenance necessary to correct faults in telephone cables
- Choose the maintenance procedure to repair equipment (locomotives, buses, industrial machinery, etc.)
- Selection of illnesses through chemotherapy treatments
- Debugging computer programs
Repair System
Systems to develop and implement plans to administer a remedy for some diagnosed problem.
- Auto repair
- Repair of electronic equipment
- Calibration of instruments
Instructional Systems
Systems developed for instruction in independent learning.
- Training people on the operation of equipment
- Development of intelligent manuals
- Troubleshooting manuals
- Instruction on specific content
Control Systems
These systems are destined to rule by controlling the overall behavior of a system. They repeatedly interpret the current situation, predict the future, diagnose the causes of the problems, formulate a remediation plan, and monitor its implementation in order to ensure the success of control. They interact with models from deterministic control theory.
- Control of production processes
- Control of operations
Benefits of Expert Systems
- Facilitates staff training
- Transfer capacity decisions
- Can be replicated indefinitely, having as many of them as required
- Solve problems that require specialized formal knowledge
- Obtain conclusions faster than human experts
- Have at least the same competence as a human expert
- Recommended where human experts are scarce