MSc: Knowledge Acquisition

  1. Introduction
  2. Course Aims and Objectives
  3. Definitions
  4. Outline of knowledge acquisition
  5. Critiques
Reading:

Cooke, N.J., 1994, Varieties of knowledge elicitation techniques, International Journal of Human-computer Studies, 41: 801-849.

Cordingly, E.S., 1989, Knowledge elicitation techniques for knowledge-based systems, In: D. Diaper, (Ed.), Knowledge Elicitation: Principles, Techniques and Applications, Ellis Horwood.

Welbank, M. A., 1990, An overview of knowledge acquisition methods, Interacting with Computers 2 (1): 83-91.

Tutor

Rob Stephens

Room 2P27

Tel: 3136

Robert.Stephens@uwe.ac.uk

Introduction

Outcome of this lecture - you should

Resources provided by tutor

Although these questions may be discussed during tutorial time, they are primarily intended to provide some form of self assessment. These questions also offer guidance of what to expect in the exam.

Learning Approaches

Assessment

Course Aims and Objectives

Practical mastery and theoretical understanding of technique

Understanding key concepts

Understanding the scope and limitations of knowledge elicitation

Get your hands dirty

Knowledge elicitation, representation and acquisition

Knowledge elicitation is those activities undertaken by a person, the knowledge elicitor, to

  1. obtain material from any relevant source,
  2. analyse and interpret that material, and
  3. put in a pre-encoded form which, while useful to those who will encode the knowledge in the KBS language, also allows it to be scrutinised by all parties in KBS development

Source: Cordingly

Knowledge acquisition involves the following:

  1. Employing a technique to elicit data (usually verbal) from the expert.
  2. Interpreting these verbal data (more or less skillfully) in order to infer what might be the expert's underlying knowledge and reasoning process.
  3. Using this interpretation to guide the construction of some model or language that describes (more or less accurately) the expert's knowledge and performance.
  4. Interpretation of further data is guided in turn by this evolving model.
  5. The principle focus for the knowledge acquisition team should be in constructing models, in domain definition, or problem identification and problem analysis.

Source: Kidd

Knowledge representation embraces following:

  1. Knowledge representation is a useful way to make explicit the knowledge that resides in an organization.
  2. Knowledge representations are constructs intended to capture and organize knowledge, both tacit and explicit, that pertains to someevent or process.
  3. The roots of knowledge representation lie in the software engineering field, wehre it is defiend as the symbolic structures which represent knowledge and the reasoning mechanisms needed to answer questions and to assimilate new information.
  4. Knowledge representation is also a major area of research in artificial intelligence (AI) and cognitive science. Any computer system that is complex enough to perform even narrowly defined human-like tasks requires enormous amounts of background information or knowledge. Structuring this vast reservoir of information so that it can be consistently represented and quickly accessed by a computer is thegoal of the AI approach to knowledge representation.
  5. Knowledge engineering is a set of principles, methods and tools for eliciting and describing specific types of knowledge and expertise and then bringing them to bear more broadly through automated knowledge-based systems (KBS). It too uses knowledge representations in its methodologies.
  6. In knowledge management, knowledge representation is used as a tool to structure the many fragmented pieces of information relevant to a particular business process or event which exist inone person or among several people. Representation shemes include maps, books of experts, and electronic 'webs'. These representations are not repositories of information; instead they are used as learning enablers, for knowledge sharing and communication.

Source: Gierkink and Ruggles

Background to Knowledge Acquisition

Computer Science

Mathematics

Linguistics

Psychology

Philosophy

Economics and Management Science

Knowledge elicitation Methods

These tend to fall into one of the four following categories:

  1. Interview methods, extended to include not just structured, tutorial, teachback and focussed interviews, specific techniques such as goal decomposition, intermediate reasoning steps, introspection and retrospection;
  2. Observation methods, including participant observation and simulation;
  3. Multi-dimensional techniques such as card sorting, repertory grid, multidimensional scaling and proximity analysis; and
  4. Protocol analysis, or self-report methods.

Metaphorical evolution of knowledge acquisition

Knowledge

Roles for knowledge acquisition

Knowledge engineering and management

technological innovation

ontology construction

document mark-up

AI systems development

generic methodologies (eg KADS)

Organizational analysis

process approaches

Task analysis

job design

User analysis

generation of cognitive specifications for tasks,

the mitigation of human error in domains of risk or time pressure

the enhancement of proficiency through training and skill remediation

Requirements elicitation

systems or design analysis

conceptual database design

software requirements definition

Types of knowledge representation

tacit

intermediate or mediating

formal or encoded

Critics of the Project of AI and KBS

From within computing

Weizenbaum

Winograd and Flores

Brookes and 'non-representational' AI

Alison Adam

From philosophy

Searle

Dreyfus

Button, Coulter, Lee and Sharrock
Ordinary Language Philosophy

Other critiques

Situated Anthropology

Sociology

Tutorial

Rob Stephens