Expert Systems (ES): Features, Classification, and Limitations

Expert Systems (ES): The expert system is one of the most active and extensive topics in artificial intelligence (ai) application research. In AI (artificial intelligence), It is a computer system that emulates the decision-making ability of a human expert. Since the first expert system DENDRAL was introduced at Stanford University in the United States in 1965, after 20 years of research and development, by the mid-1980s; various they have spread across various professional fields and achieved great success. Now, they are more widely used and further developed in application development.

Here explains Expert Systems (ES) in their points of meaning, Features, Classification, Development, and Limitations.

Expert systems are a component of artificial intelligence (ai); which contains a large amount of expert-level knowledge and experience in a certain field, and can use human expert knowledge and problem-solving methods to deal with problems in this field. That is to say, it is a program system with a lot of expertise and experience. Also, It uses artificial intelligence and computer technology to make inferences and judgments based on the knowledge and experience provided by one or more experts in a certain field Decision-making process to solve complex problems that require human experts to deal with.

In short, It is a computer program system that simulates human experts to solve domain problems. In 1982, the Japanese Government made an ambitious plan to build fifth-generation computers. This project aimed to use the computer for the following objectives i.e. Conversation, see objects, and adapt to new tasks; it will have memory and reasoning capabilities.

Inspiring with these projects both the US and European governments initiated and directed their resources in the direction to develop advance computer systems. Within a short period, AI and ES (Artificial Intelligence and Expert System) became the national concern as critics suggested that the Japanese were intent on dominating the information industry of the 1990s and beyond.

Features or characteristics of the expert systems (ES):

In general, the expert systems (es) has some common characteristics and advantages, and its characteristics mainly have three aspects in artificial intelligence (ai):

1] Inspiring:

They can use expert knowledge and experience to make inferences, judgments, and decisions. As well as, most of the work and knowledge in the world are non-mathematical. Only a small part of human activities are based on mathematical formulas or numerical calculations (about 8%). Even in the chemistry and physics disciplines, most of them are based on reasoning. Thinking; the same is true for biology, most medicine, and all laws. Also, thinking of enterprise management is almost entirely based on symbolic reasoning, rather than numerical calculation.

2] Transparency:

They can explain its reasoning process and answer the questions raised by the user so that the user can understand the reasoning process and improve the trust in them. For example, if a medical diagnosis expert system diagnoses a patient with pneumonia and must be treated with an antibiotic; then this expert system will explain to the patient why he has pneumonia and must be treated with an antibiotic; just like A medical expert explained the patient’s condition and treatment plan in detail.

3] Flexibility:

They can continually increase knowledge, modify the original knowledge, and continuously update. Because of this feature, they have a very wide range of applications. Also, expert systems with application ranking.

Classification of Expert System (ES):

The expert systems used in a specific field can be divided into the following classification or categories in artificial intelligence (ai):

1] Diagnosis:

Based on the observation and analysis of symptoms, deducing the cause of symptoms and troubleshooting methods, such as medical, mechanical, economic, etc.

2] Interpretation:

A type of system that interprets deep structures or internal conditions based on surface information, such as geological structure analysis, material chemical structure analysis, etc.

3] Predictive:

A type of system that predicts the future situation based on the current situation, such as weather forecast, population forecast, hydrological forecast, economic situation forecast, etc.

4] Design:

A type of system that designs products according to the given product requirements, such as architectural design, mechanical product design, etc.

5] Decision-making:

A type of them that comprehensively evaluates and optimizes feasible solutions.

6] Planning:

A type of them used to formulate action plans, such as automatic program design, military plan formulation, etc.

7] Teaching:

A type of them that can assist in teaching.

8] Mathematical:

A type of them for automatically solving certain mathematical problems.

9] Surveillance:

A type of them that monitors certain types of behavior and intervenes when necessary, such as airport surveillance, forest surveillance, etc.

Development of Expert Systems (ES):

There are indeed some limitations in the development of the current expert system. In the coming years, many of today’s expert system deficiencies will be improved. I believe that the projects that they should continue to study in the future include: As well as, the ability to deal with common sense; the development of deep inference systems; Also, the ability to explain at different levels; the ability of expert systems to learn; the distributed expert system; As well as, the ability to easily acquire and update knowledge.

The future development of they can directly receive data from the outside world through sensors, and can also obtain data from the knowledge base outside the system. In addition to inference in the inference machine, it can draw up plans and simulate problem conditions. The knowledge base contains not only static inference rules and facts, but also dynamic knowledge such as planning, classification, structural patterns, and behavior patterns.

Advantages of the expert systems (ES):

In the past 20 years, they have developed rapidly, the application field is getting wider and wider, and the ability to solve practical problems is becoming stronger and stronger. This is determined by the excellent performance of the expert system and its important role in the national economy.

Specifically, the advantages of expert systems include the following in artificial intelligence (ai):

• They can work efficiently, accurately, thoughtfully, quickly, and tirelessly.
• It is not affected by the surrounding environment when solving practical problems, and it is impossible to miss or forget.
• It can make the expertise of the experts not limited by time and space, to promote the precious and scarce expert knowledge and experience.
• They can promote the development of various fields, enables the professional knowledge and experience of experts in various fields to be summarized and refined, and can widely and effectively spread the knowledge, experience, and ability of experts.
• They can gather and integrate the knowledge and experience of experts in multiple fields and their ability to collaborate to solve major problems. It has more profound knowledge, a richer experience, and stronger working ability.
• The level of the military expert system is one of the important symbols of a country’s national defense modernization and national defense capabilities.
• The development and application of the expert system have huge economic and social benefits.
• Research can promote the development of the entire science and technology.

Limitations of Expert Systems (ES):

Limitations, problems, and demerits of an expert systems (es) are as follows in artificial intelligence (ai):

• It is hard, even for a highly skilled expert to abstract good situational assessment when he is under time pressure.
• The expert system performs well with specific types of operational and analytical tasks.
• The designing and construction of an expert system require expert engineers, they are rare and expensive. This limitation makes an expert system very costly.
• It excels only in solving specified types of problems in a limited domain of knowledge.
• The vocabulary that experts use for expressing facts & relations is frequently limited.
• Another limitation is that most experts have no independent means of checking whether these conclusions are reasonable or not.
• The approach of each expert to the assessment of the situation may be different, yet it may be correct.
• The expert system is comparatively costly to develop and maintain.