Expert System Concept

2.1 Definition of an expert system 
 In computer science , many experts are concentrating on the development of artificial intelligence or artificial intelligence ( AI ) . AI is a specialized study in which the goal is to make computers think and act like humans . Many AI implementations in the field of computers , for example, the Decision Support System ( Decision Support System ) , Robotic , Lanbguage Natural ( Natural Language ) , Neural Network ( Neural Network ) , and others .Examples of other areas is the development of artificial intelligence expert system that combines knowledge and data retrieval to solve the problems that normally require human expertise . Where an expert system is a form of an effort to preserve or extend the expertise of an expert in the form of computer software so that skills are not lost / extinct . This is so the expertise of an expert in a field can be extended to insert into a database and used by the public. Besides, this system can be a smart consultant or advisor in an environment specific expertise , as a result the set of knowledge that has been collected from several experts . Thus even a layman can use this system to solve the problem at hand .2.2 The characteristics of expert systems 
There are a variety of features and characteristics that distinguishes an expert system with other systems . These traits and characteristics of the main reference in the development of expert systems . Traits and characteristics are as follows :1 . Sciences expert system is a concept , not a numerical form . This is because the computer doing the processing of numerical data , while the expertise of an expert is the facts and the rules , rather than numerical .2 . Information in expert systems are not always complete , subjective, inconsistent , and subject to constantly changing depending on the environmental conditions so that decisions are uncertain and not absolute " yes " or "no " but according to a certain measure of truth . Therefore dubutuhkan system's ability to learn independently in resolving problems with special considerations .3 . Possible in an expert system to a problem are varied and have many options acceptable answer , all the factors that have traced the problem space is wide and uncertain . Therefore we need the flexibility of the system in dealing with possible solutions to the various problems .4 . Changes or the development of knowledge in an expert system can occur at any time even during the time that is needed to easily modify the system to accommodate the amount of knowledge that the bigger and more varied .5 . Views and opinions of each expert is not always the same , which is why there is no assurance that the expert system solution is surely correct answer . Each expert will membeikan pertimbangna - judgment based on objective factors .6 . Decision is the most important part of an expert system . Expert system should provide accurate solutions based on input pegetahuan though the solution is difficult to facility information system should always be required .2.3 The fields of expert system developmentThere are many areas or work areas into areas of AI , namely neural networks , perception systems , robotics , scientific language , decision support systems , and management information system based expert system .Each work area AI has the potential to solve the problem , but there are major advantages in expert systems . Heuristic itself comes from the Greek meaning Eureka find . Heuristics in expert systems do not guarantee results semutlak other artificial intelligence systems , but offers specific results to be used as an expert system to function consistently as a human expert , offers advice to users and to find solutions to specific problems .There are various categories of development of expert systems , among others :1 . Control
 Examples of development is found in the case of a patient in a hospital , in which the expert system 's ability to do control the means of treatment and care through a data sensor or alarm code and provide appropriate solutions for patients in the hospital .2 . DesignsExamples of expert systems in this field was made ​​by Dincbas PEACE in 1980 to assist the development of the design of electronic circuits . Another example is the expert system to assist with the design of computer components.3 . Diagnosis
  Development of an expert system is the largest in the field of diagnosis , such as disease diagnosis , the diagnosis of motor vehicle engine failure , malfunction diagnosis of computer components , and others .4 . Instruction
 Instruction is the development of an expert system that is very useful in the field of science and education , where the expert system can provide specific instruction and teaching of a topic issues . Examples of the development of expert systems in this field is an expert system for teaching the English language , an expert system for teaching astronomy , and others .5 . Interpretation
 Expert system developed in the field of interpretation of the process of understanding a situation of some of the recorded information . Examples of the system was developed today is the system to perform image and sound sensors then analyze it and then make a recommendation based on the recording.6 . Monitors
  Expert system is widely used in the military , which uses a radar sensor and then menganalisnya obyej determine the position based on the position of the radar .7 . Planning
 Planning is widely used in the fields of business and finance a project , where the expert system in a work plan based on the amount of labor , cost and time so that the work becomes more efficient and more optimal .8 . PredictionThis expert system is able to predict future events based on information and models of problems faced . Systems typically provide a simulation of future events , such as predicting the degree of crop damage if attacked by pests within a specified period . The program was created in 1983 by the name Boulanger PLANT .9 . Selections
 An expert system to identify the best choice of selection of multiple choice list of possible solutions. Usually the system identifies specific problems , then try to find a solution that comes closest to the truth .10 . Simulation
 This system processes the operation of several variations of existing conditions and displays them in the form of simulation . An example is a program that already incorporates PLANT between prediction and simulation , in which the program is capable of analyzing a wide range of pests with temperature and weather conditions .Programming languages ​​also helped shape the development of expert systems in the areas mentioned above . In the 1970's which is still text-based operating systems , the development of expert systems only use programming languages ​​such as prolog , LISP or Shell so that the development of expert systems at the time it becomes very difficult . The difficulty factor is also influenced by the speed of the processor and the memory is still very limited so that the expert system can only be developed on a workstation computers .In contrast to the development of computers in the early years of AI evolve , today's computers have had a gigahertz speeds and supported by the development of hardware and software that is very sophisticated .Very interesting development is the development of object -oriented programming Visual database that can process the very large amount of data . Extensively to the development of systems based on artificial intelligence , including expert systems .2.4 Advantages and disadvantages of expert systemsIn a system certainly has its advantages and disadvantages. Below are described some of the advantages and disadvantages of expert systems .2.4.1 AdvantagesThere are many benefits or advantages that can be obtained by developing expert systems , among others :1 . Ordinary people can take advantage of non - experts in the particular field of expertise without the immediate presence of an expert .2 . Increased productivity , which increases the efficiency of a particular job and the results of a working solution .3 . Savings time in solving complex problems .4 . Provide simplification solutions for complex cases and over and over again .5 . Knowledge of an expert can be documented without any time limit .6 . Allows the incorporation of various fields of knowledge of various experts to be combined .Here is a comparison between the ability of human experts and computer experts into consideration the development of expert systems .Human Experts Computer ExpertsLimited time because people need a break . Not limited because it can be used at any time .The place is local access at a single place where the experts are . Can be used in various places .Knowledge is variable and may change depending on the situation . Knowledge besifat consistent .Speed ​​varied to find its solution . Speed ​​to provide a consistent solution and faster than humans .Costs to be paid for the consultation is usually very expensive . Costs were cheap .

2.4.2 LossesIn addition to many benefits , there are also disadvantages of expert system development , namely :1 . Power and productivity of human labor is reduced because everything is done automatically by the system .2 . Development of expert system software is more difficult than with conventional software . It can be seen from the following comparison table :Conventional Software Expert System SoftwareFocus on Focus on the problem solutionDevelopment can be done individually . The development work carried out by the team .Development of conventional iterative development .2.5 Stages of Development of expert systemThere are six phases or stages in the development of expert systems . The following explanation is an explanation of an outline of the development phases :1 . IdentificationThis stage is the stage of determining important matters as the basis of the problem to be analyzed . This stage is the stage to assess and limit the problem to be implemented in the system . Any problems will be identified to look for a solution , the facilities will be developed , determining the type of programming language and objectives of the development process . If the problem identification process is done properly it will achieve optimal results .2 . ConceptualizationThe results of the identification problem is conceptualized in the form of relationships among the data , the relationship between knowledge and important concepts dam ideal to be applied in the system . Conceptualization also analyzed the critical data that should be explored bersamadengan expert in the matter . This is done to obtain confirmation of the results of interviews and observations so that the results can give a definitive answer that target the right problems , true and appropriate .3 . FormalizationIf the conceptualization stage has been completed, then in the formalization stage concepts are formally implemented , for example, provide a category system to be built , paying particular attention to several factors such as the skills of human decision-making , the difficulty and the level of difficulty that may occur , job documentation , and so on .4 . ImplementationWhen fully formalized knowledge suah , the implementation phase can already start by making an outline of the problem and then solve the problem into modules . To facilitate it should be identified :a. What are the inputb . What is the process described in the flow chart and the rule basec . What are the outputs or results and conclusionsAfter that everything changed in a language easily understood by a computer using phases such as picture stages of expert system development phase .5 . EvaluationExpert systems are completed , need to be evaluated to test and find faults . It is a common practice because of a system is not necessarily perfect after completion of manufacture so that the evaluation process required for perfecting. In the evaluation will be found the parts that must be corrected to match the problems and the ultimate goal of making systems .6 . Development systemDevelopment of the system is needed so that the system is not built to be obsolete and investment system is not in vain . It 's most useful development system is a system in which the documentation process in which stored all very important things that can be a measure of system development in the future including the knowledge dictionary problem solved .2.6 Components of an expert systemThe components of the expert system consists of several parts including : Knowledge Base ( Base knowlwdge ) , Database ( Data Base ) , Inference Engine ( Inference Engine ) , Interface ( Interface ) .2.6.1 Knowledge Base ( Knowledge Base )The knowledge base is the core program in which the expert system knowledge base is a representation of an expert's knowledge . The representation of knowledge such as facts , rules or procedures as well as heuristic knowledge available in the system . The design of the knowledge representation form unference affect engine design process , updating knowledge , and overall system efficiency .2.6.2 DatabaseThe database contains all the parts that go round the facts , good facts early on when the system began operating and the facts are taken at the time when the process is running . In effect many expert systems contain a database to store the results of the research and the data required for penelolaan linnya .
2.6.3 Inference EngineInference engine is the part that contains the mechanism of function of thinking and reasoning patterns of the system used by an expert . This mechanism will analyze a problem and will seek the best answers or conclusions . Inference engine has two functions , namely inference and control . Inference is the process of reasoning , while the control function controls the execution . Inference involves the process of watching ( and unifaction matching ( merging ) . Process is based on a database containing facts , usually stored in a special file and can also be obtained from consultations and used in the testing process implied rules of the knowledge base . two inference techniques , namely :1 . Tracking backward ( backward chaining )2 . Tracking forward ( forward chaining )Tracking the future is a necessity and backward tracking which start reasoning from a set of data leads to a conclusion .OBSERVATIONSRULES OF A FACT 1OBSERVATION RULES DRULE B 2 OBJECTIVE FACTSOBSERVATION RULES ERULE C FACT 3OBSERVATIONSFigure 2.1 Diagram of tracing backward ( backward chaining )
  RULE C CONCLUSION
                                                 FACT 1
                     RULES OF A RULE D CONCLUSIONFACT 2 OBSERVATIONS
                     CONCLUSION E B RULES RULESOBSERVATIONS F FACT 3
 CONCLUSIONFigure 2.2 Diagram of tracing forward ( forward chaining )2.6.4 User Interface ( User Interface )Between mica users are part of the link between the program and the expert system users here will be a dialogue between the user and the program . The program will ask the questions in the form " yes or no " or a form of menu choices .User interface they include:1 . Controls the display2 . Equipment input ( keyboard , mouse , etc. )3 . Controls dialog4 . Facilities assistance , explanations , suggestions5 . Models of interaction6 . Explanation of question2.7 Knowledge RepresentationKnowledge is the basis of the ability of an expert system , to realize the existing capabilities . It must however be able to present this knowledge into a form method and then used to support the process of reasoning in expert systems .2.7.1 Types of knowledgeBased on the source of knowledge can be divided into formal knowledge ( deef knowledge) and non-formal knowledge (shallow / surface knowledge) . Meanwhile, by way of representing knowledge is divided into heuristic knowledge , procedural knowledge and declarative knowledge .1 . The Formal KnowledgeFormal knowledge is knowledge that is contained in the journals , scientific books and so forth bulletin , and is often regarded as general knowledge .2 . Knowledge of Non -FormalNon-formal knowledge is the knowledge possessed by a person based on his experience in the field is a relatively long period of time .3 . Heuristic KnowledgeHeuristic knowledge is shaped hirarko knowledge , this knowledge is usually shaped tree of knowledge .4 . Procedural KnowledgeProcedural knowledge is knowledge that can be represented as a procedural process , and is stored in coded form . In essence a form of procedural programming algorithm because it contains information on how to run a particular job .5 . Declarative KnowledgeDeclarative knowledge is knowledge that can be stored in the form of data files , so that the storage can be separated . Declarative knowledge is structured based on rules and facts . ( Azis , 1994)2.8 Method of Knowledge RepresentationIn knowledge representation in expert systems , there are several methods: Method Predicate Calculus , Frame ( Frame ) , Semantic Networks ( Semantic Network ) , Methods of Production Rule .2.8.1 Methods of CalculusPredicate calculus is a simple way to represent knowledge declaratively . In predicate calculus , a declarative statement that is divided into two parts predicate and argument .2.8.2 Frame ( frame )The frame is a block or pieces that contains knowledge about specific objects , locations of events or situations or other elements with relatively large size ynag . These blocks describe the object in more detail .2.8.3 Semantic Networks ( Semantic Network )Semantic network is a way of presenting knowledge of the oldest and most straightforward . This method is a graphical depiction of knowledge that shows the hierarchical relationships of objects , the object is a node in a graph and relationships between objects represented by connecting lines labeled .2.8.4 Production RuleThe method of production rules is usually written in the form of if - then ( IF - THEN ) . This rule can be considered as relations implications of two parts, namely the premise ( if ) and the conclusion section ( then ) .2.9 Tool expert system developmentTools to build an expert system consists of two parts: the programming language and shell .2.9.1 Programming LanguageIn general, all computer programming languages ​​can basically be used for an application program to be created , but simply use without knowing the advantages and disadvantages in making expert system program , there will be an error in the program will be made . From the above problem it is necessary to be able to know the workings of the programming language .Visual Basic 6.0 programming language . This program is a visual programming language which is suitable and in accordance with the wishes of the current users generally want products that are easy to use and attractive appearance . The software is also able to connect Dbase and Paradox as well as other applications that support .2.9.2 ShellShell is a program of expert system knowledge base is empty. There are five types of shell based on knowledge representation methods used are: 
1 . Simple Rule - Base ToolsThis shell type using IF - THEN representing knowledge in this tool can be run on a personal computer ( PC ) and can manage up to 500 rules example of this tool is EXSYS , IN SIGH T2X , VP EXPERT , and ESP ADVISOR .2 . Inductive ToolsThis shell type generate rules from examples in the database pengethuan This tool is divided into two types , namely large inductive run on the frame and a small inductive running on the PC .3 . Structured Rule Base ToolsThis shell type using IF - THEN rules are compiled in to represent knowledge .4 . Logic Base ToolsShell uses the terms hom and resolution is determined by the purpose of the predicate calculus , most of these tools are built with a prologue in representing knowledge .5 . Frame Base ToolsThis shell is often described as a general object-oriented infuse each knowledge representation techniques , including the frame in one package .