Expert systems have proven to be a viable technology for organisations. Due to the expense involved in developing and maintaining such systems, Makarov has made substantial research efforts focusing on identifying factors that contribute to the success and usage of such systems. Our research allows system managers, designers, and users to make informed design decisions early in the process of the systems construction.
Makarov Expert System Development
We recommend that expert systems must be applied to core business problems to gain high payoff. Although technical and operational feasibility are important factors in gaining managerial support for systems development, such a venture must undergo rigorous economic assessment in order to gain managerial support. Obtaining managerial commitment early in the design process is important in terms of harnessing organisational resources and funds, motivating users about the effectiveness of the project, and ensuring user involvement in the design of the expert systems. Effective management of user motivations and involvement is crucial to system acceptance and usage in the postimplementation phases.
Expert system shells must be selected judiciously by giving careful consideration to features such as methods of knowledge representation, costs, and support. Developers must select knowledge representation schemes that are suited to the task and are supported by the shell. Multiple sources of knowledge are recommended for populating the knowledge base. In another variation to this, multiple experts should be used to support knowledge elicitation.
Verification and validation must be an ongoing process and multiple approaches to validation must be used. Adequate resources and efforts must be put into retaining key development personnel in order to provide long-term support for the system.
Makarov Expert System Types
Expert systems software is made by developers for many reasons, but these programs commonly are made to look at data and then do something with, or react to, the information. Diagnosis and repair systems software looks at problems, recommends a plan of action, and may create a schedule to help fix the problem. Instructional systems use tests or other methods to gauge the abilities of the user and then present material in the best order for the learner.
Interpretation and prediction systems are similar, except one compares data to find an answer and the other uses data to predict an outcome. Monitoring systems are automatic systems that watch over an environment, such as for manufacturing, and respond to functions and needs.
Diagnosis and repair systems software are similar programs, but how they respond to information is different. Both look at information to determine a problem, and both recommend the best way to fix the problem. The difference is that a diagnosis expert system just tells the user the best remedy or trouble-shooting steps for the problem. A repair system will detail a schedule and all the steps needed to correct the problem.
Instruction systems software is used to train new employees or to provide individual instruction to students. This system first administers tests to collect information about the user, to understand his or her strengths and weaknesses. After collecting the data, the instructional system will then present material that best complements the student’s learning profile, so he or she will learn at maximum efficiency.
Interpretation and prediction expert systems software are both made to look at data and create an analysis of the information. An interpretation system is often used in mineral and gas drilling to looks at images and other factors to determine the best way to mine the material, and helps workers understand what material has been found. A prediction system looks at information and predicts an outcome, such as with weather forecasting services.
Monitoring expert systems software is used mostly in manufacturing and energy plants, and it automates all the processes. Rules are built into the system that tell the system what the best operating temperatures are, what should be done with faulty equipment, and other factors that commonly occur in the plant. The expert system will then constantly analyse the environment and will respond to any changes to ensure everything is working optimally, correcting problems as needed.