Knowledge-Based (Expert) System
Software programs that store the knowledge of a human expert and are then used as consultants in particular fields are called Expert Systems. In other words, knowledge-based (expert) systems, or simply expert systems, use human knowledge to solve problems that normally would require human intelligence. Following ore some properties of this system:
• The task performed by expert system resembles one performed by a human expert.
• An expert system is developed and maintained by a knowledge engineer and expert.
• This system relies on knowledge and rules of thumb.
• The knowledge that is incorporated into an expert system usually changes constantly. It has three major components as follows:
a. User Interface
All user inputs and system-generated outputs are handled by this software using interface. Through a keyboard, users enter inputs in the form of commands and data entry operations. As a result, expert system outputs are returned to users in the form of messages and output displays.
b. Knowledge Base
It resembles a database. Knowledge base contains knowledge just as a database contains data. However, it contains the knowledge of an expert in a certain field.
c. Inference Engine
It is an integral part of every expert system lies between the user interface and knowledge base. The engine applies the data provided by the user to the knowledge base to draw conclusions.
And the three minor components as follows:
a. Working Memory
As the engine applies the data provided by the user to the knowledge base to draw conclusions, the working memory saves that user’s provided data for instant use in future.
b. Explanation Subsystem
An explanation subsystem contains additional information regarding the user’s requests and inquiries while the expert system is executing for results. The expert system also uses this subsystem to give reason for requesting the additional information.
c. Knowledge-Acquisition Subsystem
The heart of every expert system is the knowledge being updated / reviewed by an expert to the knowledge base. This update is accomplished by a software component known as a knowledge acquisition subsystem.
Few well-known expert systems in diverse business organizations are as follows:
1. DEN DRAL examines the spectroscopic analysis of an unknown chemical compound and predicts its molecular structure.
2. MYCIN is a landmark expert system used to diagnose infectious diseases and suggest possible therapies.
3. PROSPECOTOR is a geological expert system used by geologists to help find valuable mineral deposits. It also provides consultation services regarding the investigation of sites for ore-grade deposits by generating graphics in the form of maps.
4. INTERNIST helps analyze difficult clinical problems. It models a clinician’s diagnostic reasoning for a large number of diseases and their possible combinations.
5. XCON that was developed for digital equipment ensures that the different components of a computer system chosen by its customers are compatible as one system. If it is not, the expert system recommends adjustments.
6. DELTA used to help maintenance people repair diesel-electric locomotives.