Expert System (artificial intelligence)
An expert system is a sophisticated computer program designed to mimic human reasoning and problem-solving skills by utilizing a vast knowledge base and a set of rules or algorithms. These systems are interactive, allowing users to input information and receive feedback or solutions based on their queries. Central to expert systems is their inference engine, which interprets data using established rules to provide accurate responses.
Expert systems have a range of applications across various fields, including healthcare, finance, and transportation, where they can diagnose conditions, guide investment decisions, or control complex machinery. While they offer advantages such as consistency, speed, and the ability to retain information without forgetting, expert systems also face challenges. They lack human common sense and may provide solutions that are impractical or incorrect if the underlying data or rules are flawed. As technology evolves, expert systems continue to play a crucial role in advancing artificial intelligence, with ongoing developments aimed at enhancing their capabilities and applications in real-world scenarios.
Expert System (artificial intelligence)
An expert system is a computer program that uses reasoning and knowledge to solve problems. Expert systems are usually interactive in that users input information and receive feedback or a solution based on this input. Well-designed expert systems are said to simulate human intelligence so closely that the results are similar to those that would come from a highly learned human being—an expert.
Expert systems are an important area in the field of artificial intelligence (AI). AI researchers aim to use technology to develop intelligent machines—that is, machines that can use deduction, reasoning, and knowledge to solve problems or produce answers. Expert systems are highly complex and utilize advanced technology as well as scientific ideas about thought and rationality.
Researchers are interested in creating intelligent machines in part because they can be helpful to humans. Expert systems technology can produce machines that are capable of "learning" through experience—that is, they learn as they are exposed to different kinds of input.
Most expert systems include at least the following elements:
- Knowledge base: An expert system’s knowledge base is all of the facts the system has about a subject. Expert systems are usually supplied with a huge amount of information on a particular topic. This information comes from experts, databases, electronic encyclopedias, etc.
- Rules set: The knowledge base also includes sets of rules, or algorithms. These rules tell the system how to evaluate and work with the information and how to answer or approach queries from users.
- Inference engine: An inference engine uses the knowledge base and set of rules to interpret and evaluate the facts to provide the user with a response or solution.
- User interface: The user interface is the part of the expert system that users interact with. It allows users to enter their input and shows them the results. User interfaces are designed to be intuitive, or easy to use.
Different types of expert systems exist and are used for a variety of purposes. For example, some can diagnose disease in humans while others can identify malfunctions in machinery. Still other expert systems can classify objects based on their characteristics or monitor and control processes and schedules.
Brief History
Simple expert systems have existed for decades. In the 1970s, researchers at Stanford University created an expert system that could diagnose health problems. The system was more effective than some junior doctors and performed almost as well as some medical experts. Throughout the 1970s and 1980s, different researchers continued to work on expert systems that diagnosed medical conditions. Eventually, the technology began to be applied in other areas. For example, new expert systems helped geologists to identify the best locations to drill for natural resources, while others helped financial advisers to invest funds wisely.
In modern times, expert systems continue to advance. For example, automotive companies are working toward driverless vehicles; most of these vehicles include an expert system. The expert system must make decisions about accelerating, turning, and stopping, just as a human driver would. These tasks are much more complicated than the tasks of early expert systems, but they are based on some of the same principles.
Applications
Expert systems continue to affect many different aspects of society. Businesses can benefit from expert systems because they can save money by relying on a system rather than a human. Current technologies allow expert systems to handle large amounts of data, which can be beneficial for companies that crunch numbers, such as financial companies. For example, companies such as Morgan Stanley already benefit from the use of an expert system to make decisions about investments. Similarly, transportation companies can use expert systems to operate complicated vehicles such as trains or airplanes. The auto-pilot that is installed on modern airplanes is an example of an expert system; it can make decisions about navigation more quickly than human pilots.
Advantages and Disadvantages
Using expert systems rather than human experts can have some advantages. For example, an expert system’s knowledge is permanent. The system does not forget key details, as a human might. Another advantage is that expert systems are consistent; they make similar recommendations for similar situations without the burden of human bias.
Additionally, expert systems can sometimes solve problems in less time than humans, allowing them to react more quickly than people, which can be especially useful in situations where time is of the essence. They can also be replicated with relative ease, allowing for the availability and sharing of information in multiple places.
Although expert systems have many advantages, some experts have pointed out some disadvantages, too. Today’s expert systems do not have the same "common sense" as humans. That is, the system might produce answers that cannot or should not be applied in the real world. Additionally, expert systems may not recognize that some situations have no solution.
Finally, expert systems are only as good as the people who designed them, the accuracy of their data, and the precision of the rules. Thus, an expert system might make a bad choice because it is working from incorrect or incomplete information or because its rules are illogical.
Bibliography
"Definition: Expert System." PCMag. PCMag Digital Group. Web. 9 Mar 2016. http://www.pcmag.com/encyclopedia/term/42865/expert-system
"Expert System." TechTarget. SearchHealthIT. 2014 Nov. Web. 9 Mar 2016. http://searchhealthit.techtarget.com/definition/expert-system
Joshi, Kailash. "Chapter 11: Expert Systems and Applied Artificial Intelligence." Management Information Systems, College of Business Administration. University of Missouri, St. Louis. Web. 9 Mar 2016. http://www.umsl.edu/~joshik/msis480/chapt11.htm
Reingold, Eyal and Jonathan Nightingale. "Expert Systems." Artificial Intelligence. Department of Psychology, University of Toronto. Web. 9 Mar 2016. http://psych.utoronto.ca/users/reingold/courses/ai/expert.html
Russell, Stuart and Peter Norvig. Artificial Intelligence: A Modern Approach. New York: Prentice-Hall (1995): 3-27. Web. 9 Mar 2016. http://www.cs.berkeley.edu/~russell/aima1e/chapter01.pdf