Expert System Development Stages PDF

EXPERT SYSTEM – Development Stages ESC 4506 Types of Computerized Systems  Data Processing Systems (DPS) – the situation analyzed might be the existing computer system @ functional business tasks performed @ both  Management Information Systems (MIS) – the situation analyzed invo...

EXPERT SYSTEM – Development Stages ESC 4506 Types of Computerized Systems  Data Processing Systems (DPS) – the situation analyzed might be the existing computer system @ functional business tasks performed @ both  Management Information Systems (MIS) – the situation analyzed involves the manager’s business decision-making requirements and manual decision-supporting systems + existing computer databases, reports and other information systems Types of Computerized Systems  Decision Support Systems (DSS) – the situation involves interacting with managers while making a decision, requires managers’ decision needs for information + the way they make decision & perform a job  Knowledge-based Systems (KBS) – a range of different development situation Knowledge Engineering Definition:Acquire, organize, and study the knowledge to solve the problem. Process of building ES= Knowledge Engineering Knowledge Engineering Knowledge engineering addresses the problem of building skilled computer systems, aiming first at extracting the expert’s knowledge and the organizing it in an effective implementation. Knowledge Engineering  Manipulatinglarge amount of expert, heuristic knowledge into a form that a program can use to help solve difficult problems  COMPONENTS OF EXPERT SYSTEM DEVELOPMENT EXPERTS KNOWLED KNOWLED INFEREN INTERFAC USER GE GE BASE CE E S ACQUISITIO ENGINE N LITERATURE REVIEW 11 Human Elements  Domain Expert:  Person who possesses the skill and knowledge to solve a specific problem superior to other.  Knowledge Engineer:  Person who designs, and tests an ES.  End-User:  Person who will use the system. Qualifications needed by Human elements  Domain Expert  Has expert knowledge, efficient problem-solving skills  Can communicate the knowledge  Can devote time.  Knowledge Engineer  Has Knowledge Engineering skills  Has good communication skills  Can match the problem to software and has ES programming skills  End-User  Aid in the interface design, knowledge acquisition, system development. Activity of Knowledge Engineering  The knowledge engineer practices the art of bringing the principles and tools of AI research to bear on difficult applications problems requiring experts’ knowledge for their solution. The technical issues of acquiring this knowledge, representing it, and using it appropriately to construct and explain lines-of- reasoning, are important problems in the design of knowledge-based system …the art of constructing intelligent agents is both part of and an extension of the programming art. It is the art of building complex computer programs that represent and reason with knowledge of the world. (Feigenbaum 1977) Expert system development stages 1. Task analysis 2. Knowledge acquisition process 3. Prototype development 4. Expansion and refinement 5. Verification and validation 1. Task Analysis  Also known as identification  Knowledge engineer & human expert work together to identify the problem area and define scope  Identify the participants in the development process (additional experts), determine the resources needed (time, computing facilities) and decide goals @ objectives  A small but interesting sub-problems may be identified and used to focus the knowledge acquisition process 2. Knowledge Acquisition Process  Transferringproblem-solving expertise from a knowledge source and transforming it into computer program  Most important, most difficult part and most time consuming 3. Prototype Development  Represent organize knowledge into computer  Rapid prototype approach  Use development tool 4. Expansion & Refinement  Expand and refine core knowledge 5. Verification & Validation  Also known as testing  Involves evaluating performance of the prototype program and revising it to conform to standards of excellence define by experts in the problem domain  An expert system evolves:  from simple to increasingly hard tasks  Improving incrementally the organization & representation of knowledge  Systemsfor complex tasks are generally more complicated than system for simple tasks.  KBS – computer systems in which knowledge base and inferencing mechanism are separate – are often integrated with conventional procedural computer systems in variety of ways  Eg. Financial planning system contain traditional financial simulation applications + KBS to evaluate subjective decision criteria (planning assumptions) Conclusions to ES development  Knowledge engineer familiar with expert domain being put onto computer is needed for a successful development effort  Thinking skills are requirement for success  Greater levels of interpersonal skills will be needed to extract information from experts  Less control of the development effort should be given to computer information systems technical personnel  User interface considerations will be important to success  Management, technical, skills will influence the success of the effort SKILLS USEFUL IN ES DEVELOPMENT  Thinking skills  Individual interpersonal & organizational skills  Computer expertise skills Thinking skills  Have significant impact on success in developing ES esp. when analyzing and modeling knowledge-domain situations and when transforming domain-knowledge representations into representations useful for computerization  Conceptualization – to reformulate knowledge during its acquisition, analysis & representation  Inferential reasoning – drawing inferences from facts, events or anything relevant to decision maker  Developing heuristic – rules of thumb that experts usevto help them solve problems & make decisions Individual interpersonal & organizational behavior skills  Importantduring ES development phases (acquiring knowledge, directing a development effort & dealing with users)  Interviewing skills – are needed when acquiring knowledge directly from experts & in creating user- effective system  Communication skills – are needed in all phases  Skills in overcoming emotional & psychological barriers  Leadership skills  Motivation skills Computer Expertise Skills  Extensive computer expertise is not always critical except for larger system  80% of the work is not computer related