Future of Expert Systems - Report Example

?EXPERT SYSTEMS Expert Systems Affiliation Table of Contents Table of Contents 2 3 Definition and of an Expert system 3 Disadvantages of Expert Systems 5 Benefits of Expert Systems 5 The History of Expert Systems 7 Developing an Expert System 8 Identification 9 Conceptualization 9 Formalization 9 Implementation 10 Testing 10 Early Expert Systems 11 Rule Based Expert Systems 11 Fuzzy Expert Systems 12 Neural Expert Systems 13 Today’s Expert Systems 13 Legal Expert Systems 13 Medical Expert Systems 14 Web-based Expert Systems 14 Future of Expert Systems 15 References 16 Abstract At the present, information technology is playing a significant role in every walk of life. It has offered wonderful techniques which allow organizations to carry out their daily tasks smoothly. In this scenario, this paper presents an overview of expert systems. An expert system is believed to be an information system or decision support system but actually it is different from these systems. As its name indicates, expert systems are used to perform specific tasks which involve the knowledge and skills of experts. At the present, expert systems have become very important for the majority of business organizations. In fact, they are currently being used in every walk of business organization. Expert systems are used to perform different operations in different organizations. This paper presents a detailed analysis of expert systems. This paper discusses the advantages, disadvantages and evolution of expert systems. This paper also discusses the examples of earlier and latest expert systems. Definition and Description of an Expert system Although Artificial Intelligence (AI) has been around for more than fifty years, it has been just recently that a lot of organizations all over the world are beginning to utilize AI based tools and applications to help them become more competitive in the ever-increasing competitive world. In the past few years, AI has been rapidly turning into an imperative technology and there is at this time an explosion of interest centering on this field. In fact, both industries and educational institutes are assigning more resources than ever before to AI. Basically, the artificial intelligence is a wide-ranging trend and it consists of a lot of sub domains such as game-playing systems, vision systems, computer-aided instruction, natural language translation, voice synthesis and recognition, robotics, and expert systems. Expert systems are possibly the fastest progressing sub domain of AI (Schon & Helferich, 1989). Expert systems are acknowledged as a significant subject of artificial intelligence. Basically, an expert system offers a method to collect and transform the knowledge of experts. The expert system consists of computer programs that try to be like the way people think. The history of expert systems can be traced back to over twenty years ago in the labs of Stanford University where it was first used to help make a diagnosis of infectious blood diseases. In fact, since that time expert systems have been using into almost every walk of life that involves human knowledge and judgment. In addition, the expert systems are normally based on three most important elements: the clarification generator and user interface, the inference engine, and the knowledge base. Additionally, in the beginning while using expert systems at Stanford, the knowledge base encompassed medical "rules" to demonstrate IF-THEN conditions and statements with a related confidence factor. The example of this rule can be like this, IF the patient is diagnosed with symptom A AND symptom B THEN the result/disease is X, and confidence is Y%. If the patient is diagnosed with symptom C, then this rule would not even be applicable. Moreover, the decision regarding the selection of the rules is made automatically by the inference engine (MoreBusiness, 1998). In simple words, “an expert system is an information system that collects and stores the expertise of human experts and then reproduces human reasoning and decision making”. Furthermore, an expert system can be defined as a knowledge exhaustive computer program that captures the skills or expertise of human in limited fields of knowledge. An expert system can help decision making by creating related questions and describing the causes for taking definite action (Shelly, Cashman, & Vermaat, 2005, p.729; Laudon & Laudon, 1999, p.446). Disadvantages of Expert Systems Despite a number of advantages, expert systems have some limitations as well. Some of the limitations of expert systems are outlined below: (Learners House, 2012; Liberopoulou, 2006) Solving a problem using an expert system can be more time-consuming than experts. Expert Systems are not generalized problem solvers or experts. Some of the problems cannot be resolved by an expert system. It is not for all time feasible to take out knowledge from an expert. In fact, it takes a lot of time and effort to develop an expert system. It is a very complex task to make the cost-benefit analysis of by making use of an expert system. Data and information to be collected for the expert system may not for all time be readily accessible. Different experts can have different approaches to solve a single problem. There is no technique to verify that whether the results of expert systems are logical or correct. Benefits of Expert Systems As discussed above, expert systems can be used to store useful expert knowledge and make it archival. It is a significant advantage for the organizations, especially when they feel they can lose the experts would be a considerable loss to the organization. On the other hand, making use of the expert knowledge improves employee efficiency by providing the required support to make the best decision. In addition, enhancements in consistency and performance normally come into view when expert systems give out expert advice, judgment, and clarification on demand. It is an admitted fact that expert systems can carry out extremely complicated activities and tasks and a tremendously rich knowledge-database arrangement and comfortable. In this way, they are well-matched and compatible to copy human problems and activities. Additionally, expert systems can decrease production downtime and, consequently, enhance efficiency and performance. Moreover, expert systems make easy the distribution of expertise to distant locations by making use of digital communication channels (such as the Internet). Furthermore, in many cases, continuing use of an expert system would be cheaper and more reliable than the services of a human expert (ReferenceForBusiness, 2012). According to (Sagheb-Tehrani, 1993), organizations can use expert systems to get a wide variety of advantages, for instance minimizing expenditures, organizational changes, to improve/aid decision making and improving business efficiency. The first advantage is about minimizing expenses, when expert systems replace with human expertise then the need for human resource is decreased. Additionally, the implementation of an expert system usually causes organizational changes in reporting arrangement, job content, and so on. In fact, it can also require human resources learning a wide variety of techniques for carrying out their tasks. On the other hand, decreasing job numbers because of implementation of an expert system engages some degree of organizational changes. One of the most important benefits of expert system is improving organizational efficiency. It is because of the fact that expert systems allow the organizations to make effective and timely decisions which can result in higher quality and quantity in productions. In view of the fact that expert systems are giving the assurance for improved efficiency, thus organizations must keep in mind its consequences for mankind and the features of their work atmosphere. However, it can have need of active management response, which can go deeper than just the obligation for user participation. Moreover, the innovative environment of this technology requires a deeper knowledge and understanding of its design and the environment, in which this technology will be implemented (Sagheb-Tehrani, 1993). The History of Expert Systems Expert systems are believed to be the initial truthfully successes of artificial intelligence in which a number of researchers contributed. Some of the well-known researchers who took part in the success of expert systems included Edward Feigenbaum, with the Dendral and Mycin systems, Bruce Buchanan, Randall Davis, Edward Shortliffe, Carli Scott, William vanMelle, and others. On the other hand, in France, researchers paid attention to the computerization of logic engines and reasoning. In this scenario, the French version of Prolog computer language was emerged in 1972, which led the real development over expert systems such as Mycin or Dendral. However, Prolog does not appear to be mainly easy to use language and is believed to be logic further than human logic. In the 1980s, the research and developments in the field of expert systems accelerated as expert system appeared to be a realistic technology for providing solutions of various real-world problems. At that time, the majority of universities started offering courses of expert systems as well as the majority of business organizations started adopting this latest emerging technology for the management of their business operations. It was the 1981 when IBM introduced the first computer with MS-DOS operating system. In fact, its low price and useful features attracted a large number of people and offered an innovative platform for expert systems. A large number of business organizations started to import expert systems shells from universities, renamed "generators" for the reason that they added to the shell a tool for writing rules in plain language and as a result, supposedly, facilitated them to implement expert systems without making use of any programming language or any other software. In addition, in 1986, a very useful and innovative expert system generator for computers was developed and appeared on the market. This product was based on the French academic research. This expert system generator made use of propositional logic (that is known as Zeroth order logic) in order to build and test expert systems, developing clarifications and identifying logic inconsistencies between the reasoning and facts on a knowledge base developed using traditional language rules. In addition, developments and advancements in this discipline further grown in the 1990s. In this scenario, the successful implementations of expert systems were supported by the emergence of the symbolic processing languages such as Prolog and Lisp (Academic Room, 2013). Developing an Expert System Up till now a large number of general approaches for the development of expert systems have been proposed. The development process of expert system is devided into five stages, which are outlined below: Identification Conceptualization Formalization Implementation Testing In addition, all these stages are mutually dependent and interconnected. In this scenario, an iterative development lifecycle runs until the software constantly carries out at a satisfactory level (Engel, 1993). Identification At this stage, the requirements for the development of an expert system are indentified. It is similar to the requirements analysis process that is performed in traditional software development lifecycle. At this stage of expert system development a wide variety of requirements gathering activities are carried out. These activities involve a formal requirements and task analysis to find out the system requirements, establishing the motives of ES development and the users of the system and determine of the input and output. In addition, the problems, the system stakeholders, the resources, the objectives, the expenditures and the time frame needed for the development to be obviously determined at this stage (Engel, 1993). Conceptualization Conceptualization is the second stage of expert system development, which involves all the activities related to expert system design make sure that particular associations and communications in the problem area are identified and understood. In this scenario, the key issues and entities, associations between processes and objects and control methods are identified. In fact, this stage is believed to be the initial stage of knowledge acquisition. In addition, it requires the detailed classification of the circumstances as well as understands the knowledge required for solving the problem (Engel, 1993). Formalization As name indicates, formalization refers to organization of the major ideas, information flow into formal representations and sub problems emerged during analysis and design phases. To all intents and purposes, the basic logic of the expert system programming is developed at this stage. In addition, at this stage the most of the knowledge collected during previous stages is grouped or modularized, in fact is useful to present the problem solving process graphically (Engel, 1993). Implementation At this stage, the formal knowledge collected in previous stages is coded or mapped into the structure of the implementation tool to develop working software. Additionally, this stage involves organizing inference rules, contents of knowledge structures and control strategies established in the earlier phases into an appropriate structure. In addition, knowledge engineers can make use of a wide variety of application development tools to develop working software to put in order and document information gathered during the initial phase, with the intention that implementation is accomplished at this time otherwise the explanation from the previous stages are programmed at the moment (Engel, 1993). Testing Testing is the final stage of expert system development, which involves significantly more than identifying and removing errors of language rules. Additionally, this stage involves various important aspects such as the justification of program efficiency and assessment of the effectiveness of the programming tool and authentication of individual associations. In this scenario, testing directs evaluation of ideas, redesign of demonstrations along with additional modifications (Engel, 1993). Early Expert Systems Rule Based Expert Systems Every expert system is based on some knowledge. In fact, the knowledge is believed to be the vital element in development of an expert system. Basically, this knowledge has two main parts: rules governing the business and facts. In this scenario, these rules are provided in the form of ‘if–then–else’ structure. In addition, these business rules are attained from the professionals in the field. Every business has specific rules and the rules of a business cannot be applied to another business. In this scenario, the expert system developed using one kind of business rules cannot be applied to another area. Some of the major advantages of rule-based expert systems are (Nagori & Trivedi, 2012): These expert systems are developed by taking the guidance from experts as a system will be developed when an expert will explain its action. Consequently, it is easy for the development team to develop the system. In view of the fact that knowledge base is isolated from the rule processing, hence it makes it easier for the development team to alter and update new rules as knowledge increases. Some of the disadvantages of rule base expert system s are outlined below (Nagori & Trivedi, 2012): The relationship between knowledge and rule is confusing. Hence, it makes it complicated for the expert to explain series of logical relations. In view of the fact that for every new request, an expert system will move across complete set of rules. Hence, in case if there are more than 2000, then it will take a lot of time to execute the entire process. In some difficult scenarios, an expert system will not be helpful in supporting decision making for the reason that the rules for the exclusion are not programmed into the expert systems. Fuzzy Expert Systems These expert systems are developed using the fuzzy set theory or fuzzy logic. Basically, the fuzzy logic refers to fuzziness, which stands for standardize ambiguity. In this scenario, fuzziness is used to determine the extent to which particular objects can be divided into classes or cannot be divided into different classes. In addition, fuzzy logic is used to determine a group of arithmetical rules and regulations for the representation of knowledge based on level of relationship instead of on hard membership of traditional binary logic. In the development of a fuzzy expert system one of the key steps is to determine the variables that will get fuzzy values and then build rules against those fuzzy values (Nagori & Trivedi, 2012). One of the most important advantages of fuzzy expert systems is that they are very useful in situation when it is very difficult to find out the correct value of the variable, in fact it is the most important part of decision-making (Nagori & Trivedi, 2012). Biggest disadvantage of these expert systems is that not all the objects can be divided into classes as fuzzy sets (Nagori & Trivedi, 2012). In most of the cases relative levels similar to more and less happen, which requires us to describe new rules and fuzzy set values (Nagori & Trivedi, 2012). Neural Expert Systems These expert systems are developed keeping in mind the model of human brain. Similar to a human brain a neural expert system learns, modifies its knowledge, updates, and takes out new knowledge from the available knowledge base. Some of the important advantages of neural expert systems are outlined below (Nagori & Trivedi, 2012): These expert systems are capable of generalization. These expert systems are capable of dealing with incomplete and noisy data set. These expert systems are elastic in changing situation. These expert systems can provide excellent support in situation when all predictable mechanisms for system development fail. One of the major disadvantages of neural expert systems is that they cannot offer clarification capability for the solution presented by it (Nagori & Trivedi, 2012). Today’s Expert Systems Legal Expert Systems Legal expert systems are those that provide answers or solutions of the law related questions. These questions can be asked from a lawyer. These experts systems can be used by lawyers in deciding for or preparing legal argument. One of the major advantages of these expert systems is that they provide answers of law related queries quickly (Popple, 1991). Some of the major disadvantages of legal expert systems are outlined below (Popple, 1991): They do not take into consideration the helpful features of past cases. Legal expert systems are not judgment machines. They don’t allow lawyers to take decisions. These systems can be used to prepare cases or arguments. Medical Expert Systems These expert systems are designed to support healthcare firms in judgment and treatment, and they have been used in a wide variety of medical contexts. For instance, medical practitioners, nurses, hospital doctors, operation theatre and consultants. In addition, these expert systems are also used by parents, and patients themselves. Some of the major advantages of medical expert systems are (Pemberton, 2013): These expert systems provide the answers of medical related questions quickly. Cure can be suggested without wasting time. Major disadvantages of these expert systems are (Pemberton, 2013): Their knowledge base can be out of date. These expert systems need to be regularly updated. Web-based Expert Systems Web-based expert systems have emerged as a most attractive form of expert systems for the reason that they contain the advantages of both web technology and expert system technology. Basically, the use of web services to provide the features of web-based systems supports the incorporation of these systems in web-portals. At the present, these expert systems are being used in a number of areas such as management, engineering, agriculture, medicine, tourism, education, and finance (Kumar & Mishra, 2010). Some of the advantages of these systems are outlined below: The access to the Internet is widely available. We-browsers provide a common multimedia interface to support these expert systems. At the present, a large number of Internet-compatible tools for the development of web-based expert systems are available. Web-based services and functionalities are naturally compatible and portable Some of the disadvantages of these expert systems are: These expert systems require quick technological modifications and upgrading to servers, inference engines, interface components and a variety of protocols. The potential delivery of services is affected because of communication loads and a limited technological infrastructure (Duan, Edwards, & Xu, 2004). Future of Expert Systems The research has shown that the expert systems are currently playing significant roles in every walk of life. These innovative systems are being implemented and used in every industry. At the present, a wide variety of expert systems are being used in banking, business, and healthcare and so on. The growth of these expert systems is further supported by the latest tools and technologies such as the iPhone and laptops have made knowledge more accessible along the way. In addition, expert systems are capable of providing the reason about a knowledge base which is invented of an excess of rules and facts, typically associated to a particular problem area (Franz, 2012). Certainly, in the future, expert systems will play role in every area of life. In fact, the majority of expert systems in business organizations would be repositories of facts and deals, as well as several expert systems in government sector would play significant role in national security. In addition, their implementation would be a precious sign for a legal or competitor’s grievance-point for human resources or customers. Additionally, it is expected that the world will soon be controlled by expert systems. Without a doubt, expert systems have been proved to be excellent technology; however they take a lot of time and data to assemble. Though, traditional artificial intelligence ensures the development of a self-training expert system, however it has not delivered up till now. Moreover, the combination of the two would bring key changes in the world (123HelpMe, 2013). References 123HelpMe. (2013). Expert Systems: The Past, Present and Future of Knowledge-based Systems. Retrieved July 10, 2013, from http://www.123helpme.com/view.asp?id=21959 Abraham, A. (2010). Rule-based Expert Systems. Retrieved July 10, 2013, from http://www.bhu.ac.in/ComputerScience/vivek/softcomp/fuzzy_chapter.pdf Academic Room. (2013). Expert Systems. Retrieved July 08, 2013, from http://www.academicroom.com/topics/expert-systems-with-applications Binsted, K. (2009). Rule-based Expert Systems. Retrieved July 10, 2013, from http://www2.hawaii.edu/~binsted/ics661/Rule-based.pdf Duan, Y., Edwards, J. S., & Xu, M. X. (2004, August 21). Web-based Expert Systems: Benefits and Challenges. Retrieved July 10, 2013, from http://eprints.aston.ac.uk/2857/1/Web-ES-IM-published_version.pdf Engel, B. (1993). Building Expert Systems. Retrieved July 12, 2013, from https://engineering.purdue.edu/~engelb/abe565/es.htm Franz. (2012, March 10). Rise of the Expert Systems. Retrieved July 12, 2013, from http://taptanium.com/blog/posts/54/rise-of-the-expert-systems Kumar, S., & Mishra, R. B. (2010). Web-based expert systems and services. The Knowledge Engineering Review, Volume 25 Issue 02, pp. 167-198. Laudon, K. C., & Laudon, J. P. (1999). Management Information Systems, Sixth Edition (6th ed.). New Jersey: Prentice Hall. Learners House. (2012). Disadvantages of Expert Systems. Retrieved April 30, 2012, from http://learnershouse.com/disadvantages-of-expert-systems/ Liberopoulou, L. (2006). The Use of Expert Systems in Conservation. Retrieved May 03, 2012, from http://radio-weblogs.com/0101842/stories/2003/06/01/theUseOfExpertSystemsInConservation.html MoreBusiness. (1998). The Impact of Expert Systems. Retrieved April 29, 2012, from http://www.morebusiness.com/running_your_business/businessbits/v2n8.brc Nagori, V., & Trivedi, B. (2012). Which type of Expert system – Rule Base, Fuzzy or Neural is most suited for evaluating motivational strategies on human resources :- An analytical case study. International Journal of Business Research and Management (IJBRM), Volume 3 Issue 5 , 249-254. Pemberton, L. (2013). CS237 Intelligent Sytems. Retrieved July 10, 2013, from http://www.it.bton.ac.uk/staff/lp22/CS237/CS237MedicalXSys.html Popple, J. (1991). Legal Expert Systems: The Inadequacy of a Rule-Based Approach. Australian Computer Journal, Volume 23 Issue 1, pp. 11-16. ReferenceForBusiness. (2012). EXPERT SYSTEMS. Retrieved April 30, 2012, from http://www.referenceforbusiness.com/encyclopedia/Ent-Fac/Expert-Systems.html Sagheb-Tehrani, M. (1993). The technology of expert systems: some social impacts. ACM SIGCAS Computers and Society, Volume 23 Issue 1/2, pp.15-20. Shelly, Cashman, & Vermaat. (2005). Discovering Computers 2005. Boston: Thomson Course Technology. Read More