Get Started. It's Free
or sign up with your email address
Rocket clouds
WHAT I HAVE LEARNT (57033) by Mind Map: WHAT I HAVE LEARNT (57033)

1. Overview of Expert System

1.1. Artificial Intelligence

1.1.1. Intelligence performed by machine

1.1.2. designated to imitate human capabilities Thinking Sensing

1.2. Expert System (ES)

1.2.1. A computer system that emulates the decision-making ability of a human expert.

1.2.2. Knowledge Collection of information used to solve problems Such as facts, concepts, procedures, models, heuristics

1.3. Component of Expert System

1.3.1. Knowledge Base Knowledge for understanding, formulate, and solve problems eg: Facts and rules Primary raw material of ES

1.3.2. Inference Engine "Brain" of ES Based on the use of rules

1.3.3. User Interface Interaction among ES and users Language processor eg: NLP

1.3.4. Knowledge Acquisition Accumulation, transfer, and transformation of problem-solving expertise from experts Documented knowledge sources to computer program Requires Knowledge Engineer (KE)

1.3.5. Knowledge Representation (KR) Express information about the world in a form that can be utilize by computer system If...., then.....

1.3.6. Experts Experienced practioner Make good decisions Make quick decisions

1.4. Why ES/KBS

1.4.1. To achieve an expert knowledge

1.4.2. To disseminate knowledge

1.4.3. To ensure uniformity of advice or decisions

1.4.4. As basis for training other specialists

1.4.5. Advantages Economical Availability Short Response time Reliability Explanation Intellectual property

1.4.6. Disadvantages High cost Time consuming High failure rate

1.5. Application of ES

1.5.1. DENDRAl for chemistry field

1.5.2. MYCIN for biology, specialized in bacteria and antibiotics

1.5.3. PROSPECTOR for geological field

1.5.4. XCON/R I For computer system

1.5.5. Medical diagnosis

1.5.6. Strategy games Eg: Chess

1.5.7. Financial advice

2. Expert System Development Life Cycles

2.1. ES Development Team

2.1.1. Domain Expert Person with knowledge and skills of solving problem in a specific area Have major expertise in given domain

2.1.2. Project Manager

2.1.3. Knowledge Engineer Able to design, build, and test ES establishes reasoning methods being used by expert

2.1.4. Programmer

2.1.5. End user

2.2. ES Methodology

2.2.1. Cyclical Development Rapid Application Development (RAD) Approach Knowledge Acquisition-Prototype Development-Prototype Critiquing Advantages: show early progress in Knowledge elicitation task generates enthusiasm in domain expert

2.2.2. Turban Expert System Development Life Cycle Phase 1: Project Initialisation Phase 2: System Analysis & Design Phase 3: Prototyping Phase 4: System Development Phase 5: Implementation Phase 6: Post-Implementation

2.2.3. Knowledge Acquisition and Documentation Structuring (KADS) Planning Knowledge Definition Knowledge Design Code & Checkout Knowledge Verification System Evaluation

3. Knowledge Acquisition

3.1. Scope & Types of Knowledge

3.1.1. Documented knowledge

3.1.2. Undocumented knowledge

3.1.3. Levels of Knowledge Shallow Deep

3.1.4. Categories of Knowledge Declarative Knowledge Procedural Knowledge Meta Knowledge

3.2. Process of Knowledge Acquisition

3.2.1. Aims and objectives understood by KE

3.2.2. KE develops working knowledge

3.2.3. KE interacts with experts Interviews Structured Focused Meetings

3.2.4. KE produces document knowledge base interview transcript analysis of information full description of major domain entities

3.3. Knowledge Acquisition Modelling Methods

3.3.1. Semiautomatic little or no help from KE minimal participant of experts Knowledge acquisition can be supported by computer-based tools PROTEGE-II KSM

3.3.2. Automatic roles of both expert and KE are minimize and eliminated Knowledge Discovery Reasons Good KE are highly paid Difficult to find good KE Domain experts always busy Uncooperative experts

3.3.3. Manual Interviews Structured Focused Observation Discussions with users Commentaries Brainstorming Prototyping Conceptual graphs and models Case Analysis Critical incident analysis

3.4. Process Tracking Methods

3.4.1. a set of techniques attempt to track the reasoning process of expert

3.4.2. Protocol Analysis KE acquires detailed knowledge from experts Expert is asked to think aloud while performing task