Expert systems

Expert systems are artificial intelligence systems built on the basis of deep specialized knowledge of some subject areas (obtained from expert experts in these areas). Expert systems are one of the few types of artificial intelligence systems that have become widespread and have found practical application. There are expert systems for military affairs, geology, engineering, computer science, space technology, mathematics, medicine, meteorology, industry, agriculture, management, physics, chemistry, electronics, jurisprudence, etc. And only that expert systems remain very complex, expensive, and most importantly, highly specialized programs, restrains them even more widespread.

Expert systems are artificial intelligence systems built on the basis of deep specialized knowledge of some subject areas (obtained from expert experts in these areas)

Features of expert systems:

  • competence - in a specific subject area, the expert system should reach the same level as the human experts; At the same time, it should use the same heuristic methods, and also deeply and broadly reflect the subject area;

  • symbolic reasoning - the knowledge on which expert systems are based represent symbolically the concepts of the real world, reasoning also occurs in the form of the transformation of symbol sets;

  • depth - expertise should solve serious, non-trivial tasks that differ in the complexity of knowledge that expert systems use, or the abundance of information; this does not allow us to use the full search of options as a method of solving the problem and makes us resort to heuristic, creative, informal methods;

  • self-awareness - expert systems should include a mechanism for explaining how they come to the solution of the problem.

Expert systems are created to solve various kinds of problems, but they have a similar structure. Expert systems that perform interpretation, as a rule, use information from sensors to describe the situation. For example, it could be the interpretation of the meter reading at a chemical plant to determine the state of the process.

Expert forecasting systems determine the likely consequences of given situations.

Expert systems perform diagnosis using situation descriptions, behavior characteristics, or knowledge of component design to determine the likely causes of a malfunctioning diagnosed system.

Expert systems that perform design, develop the configuration of objects, taking into account a set of limitations inherent in the problem. Examples include genetic engineering, the synthesis of complex organic molecules.

Expert systems involved in planning design actions; they determine the complete sequence of actions before they begin. Examples are the creation of a plan for applying a sequence of chemical reactions to groups of atoms for the synthesis of complex organic compounds.

Expert systems that perform surveillance compare real behavior with the expected behavior of the system. Examples include the monitoring of meter readings in nuclear reactors to detect emergencies.

Expert systems that carry out training, diagnose, "debug" and correct (correct) the behavior of the trainee. Examples include teaching students to find faults in electrical circuits, training military seamen to handle the engine on the ship, and training medical students to choose antimicrobial therapy.

Expert systems, which manage, adaptively manage the behavior of the system as a whole. Examples are the management of the production and distribution of computer systems or monitoring the status of patients with intensive care.

Consider examples of the most famous classical expert systems, from which the creation and development of this type of software began.

MYCIN is an expert system developed for medical diagnostics. In particular, it is designed to work in the field of diagnosis and treatment of blood and medical infections. The system makes the appropriate diagnosis based on the symptoms presented to it, and recommends a course of drug treatment for any of the diagnosed infections.

DENDRAL is the oldest, most developed expert system in the world. A chemist, when preparing a substance, often wants to know what its chemical structure is. There are various ways for this. First, a specialist can make certain conclusions based on his own experience. Secondly, he can investigate this substance on a spectrometer and, studying the structure of the spectral lines, clarify his initial guesses. In the most general terms, the decision-making process is as follows. The user gives the DENDRAL system some information about the substance, as well as data of spectrometry (infrared, nuclear magnetic resonance and mass spectrometry), and she, in turn, gives the diagnosis in the form of an appropriate chemical structure.

PROSPECTOR is an expert system used in the search for commercially viable mineral deposits.