Expert System: Application of Artificial Intelligence
1 week ago Rishita Raj 0
An expert system is a system which is coded with knowledge in its knowledge base domains from where it gives information about the specialized domains without any human competence.This knowledge-based system is an application of AI(artificial intelligence), as AI aims at building systems which can impart its intelligent behaviour to the world. It is a computer system in which the information is coded from which the knowledge is obtained without any human competence.The system’s knowledge is obtained from experts, journals, books, expertise, and databases.The expert system consists information from various disciplines such as geology, medicine, engineering, banking, finance, meteorology and so on.
The expert system basically has two main components:
1.Knowledgebase: After an appropriate expert knowledge is acquired, it along with some protocols are encoded in some form in the knowledge base, it is verified and t knowledge is updated at certain intervals all through the life of the system.
2.Interference engine: The interference engine responses to the user inputs by the information considering the defined protocols in the given domains stored in the knowledge base by the I/O interfaces.During this process, at first, a proper matching is done i.e the contents present in the knowledge base is compared to the user inputs, in such a case many sets are matched so now comes the selection part, in this part the appropriate set is selected considering the defined protocols and at last it is executed.
The expert system is not a new term that is introduced before a century.It has emerged from the research laboratories in some of the leading universities around the 1960s and 1970s.At that time researchers aimed to build such a system that uses expert knowledge rather than algorithms and other methods.The approach was made to design such kind of system by applying AI(artificial intelligence).
DENDRAL was an expert system that was produced during 1965s using artificial intelligence at Stanford University.It was a chemical analysis expert system which analyzed the chemical compounds and spectrographic data obtained from them.DENDRAL used heuristic knowledge from the expert chemists.This heuristic Dendral paved the away for the researchers to obtain Meta-DENDRAL.Meta-DENDRAL was a machine learning system which receives the set of inputs associated with chemical structures and proposes the set of rules from positive examples.Soon after DENDRAL developed, another expert system MYCIN was developed at Stanford University, this expert system too used AI to identify bacteria causing severe infections and blood diseases and to recommend antibiotics and therapies to the patient.The MYCIN’s knowledge was improved for several years as the new knowledge was added which at last had a success rate of 69% which was better than the performance of disease experts who were judged using came criteria.
Other early expert systems which came was PROSPECTOR developed at SRI International, which assists geologists in the mineral explorations.PROSPECTOR performs a consultation to determine the best drilling site, the best model that fits the data, the additional data required if necessary, recommendations and many more.The R1(internally called XCON) was developed in the late 1970s was probably a production ruled-based system used by the Digital Equipment Corporation to select and configure components of computer systems.XCON first went into use in 1980 in DEC’s plant in Salem, New Hampshire. It eventually had about 2500 rules. By 1986, it had processed 80,000 orders and achieved 95-98% accuracy.STEAMER, another expert system which was an interactive simulation-based training system which was developed to train engineers in the steam propulsion systems.It allows to control, manipulate a dynamic physical system.So, these introductions at early ages have motivated to broaden the depth and range to revolutionize many new expert systems with a high success rate in various new disciplines as well.
In the era of artificial intelligence and expert systems, applications can now be found in almost all disciplines.Some of them are included here:
- They are widely used in various medical diagnosis (infectious blood disease, internal medicine, pulmonary disease and many more).It can find some symptoms or signs related to some diseases.It suggests medicines, therapy, antibiotics etcetera like a doctor and in some cases, also suggests the doctor if the patient does not meet diseases criteria for new evaluation plans.
- Diagnosis of a problem in complex electronics and electromechanical systems are difficult as because of complexity in the systems, constant changes in systems both internally during hardware and software updates as well as externally by the environment, the changes at different rates due to failures, loads, traffic and many more.So, in such a situation expert system plays their role smartly.
- Planning as well as conducting Experiments and researchers in scientific disciplines such as in biology, chemistry, physics, genetics and so on.
- It has also a wide range of applications in planning curriculum of the student as well as diagnosing, assessing, repairing the behaviour of the student.
- The government sector also has applications in planning for minimal taxation, maintaining and protecting data of citizens, money-laundering cases, detecting criminals and other specified goals.
- In the banking sector, it provides reduced manpower, increases in quality, security systems in banks, reduced errors, reducing administrative costs, improved customer services and so on.
- Designing and manufacturing of very big system/objects(such as VLSI systems)i.e design of components and physical processes including high ranging conceptual designs as well as manufacturing a system from its subassemblies.
- In various areas of marketing analysis such as sales, quality, competition, feedbacks etcetera and on these basis providing suggestions(like for improving quality, quality production and many more).
- Providing financial analysis to big firms such as organization of personnel, short term and long term debts, reputation in the market, profitability, position in the market.A successful financial analysis helps a firm to reduce the risk factors and to achieve a maximum profitability.
Keeping aside the advantages that we have seen in the application of the expert systems, it has few limitations which are discussed here in this section. It is cost-effective i.e it requires a good amount of cost to set up.They are machine experts who cannot learn from its mistakes and also cannot come up with new solutions(mimics solutions by experts).It requires continuous updations which for a temporary period is useless. No introspection is possible and also various expert system lacks much-required self-analysis tools.Sometimes, the expert systems are able to make some available common sense knowledge and broad details of information which lastly results in unavoidable circumstances.