Knowledge Capture Systems: Systems that Preserve and Chapter 5 Knowledge Capture Systems: Systems that Preserve and Formalize Knowledge

Chapter Objectives To describe what are knowledge capture systems To explain how to elicit and store organizational and individual knowledge To discuss the value of organizational storytelling for knowledge capture To explain the two types of knowledge capture systems To capture knowledge in educational settings To capture tactical knowledge

What are Knowledge Capture Systems? Knowledge capture systems support process of eliciting explicit or tacit knowledge from people, artifacts, or organizational entities These systems can help capture knowledge existing either within or outside organizational boundaries, among employees, consultants, competitors, customers, suppliers, and even prior employers of the organization’s new employees Knowledge capture systems rely on mechanisms and technologies that support externalization and internalization.

Knowledge capture systems The development of models or prototypes, and the articulation of stories are some examples of mechanisms that enable externalization. Learning by observation and face-to-face meetings are some of the mechanisms that facilitate internalization Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Using Stories for Capturing Tacit Knowledge Organizational stories: “a detailed narrative of past management actions, employee interactions, or other intra- or extra-organizational events that are communicated informally within organizations” include a plot (or plan), major characters, an outcome, and an implied moral play a significant role in organizations characterized by a strong need for collaboration Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Organization stories ... Important to capture and transfer tacit knowledge stories make information more vivid, engaging, entertaining, and easily related to personal experience and because of the rich contextual details encoded in stories, they are the ideal mechanism to capture tacit knowledge Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Using Stories for Capturing Organizational Knowledge Guidelines for organizational storytelling (Dave Snowden, 1999): Stimulate the natural telling and writing of stories Rooted in anecdotal material reflective of the community in question Should not represent idealized behavior An organizational program to support storytelling should not depend on external experts for its nourishment Organizational stories are about achieving a purpose, not entertainment Be cautious of over-generalizing and forgetting the particulars Adhere to the highest ethical standards and rules

Eight steps to successful storytelling Have a clear purpose. Identify and example of successful change. Tell the truth. Say who, what, when. Trim detail. Underscore the cost of failure. End on a positive note. Invite your audience to dream. In story telling, it is not just what you say but how you say it, that will determine its success.

Using Stories for Capturing Organizational Knowledge Important considerations: Effective means of capturing and transferring tacit organizational knowledge Identify people in the organization willing to share how they learned from others Use metaphors to confront difficult organizational issues e.g. time is money Storytelling provides a natural methodology for nurturing communities because it: builds trust unlocks passion is non-hierarchical (told in the form of peer dicussion) Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Other consideration for effective story telling People must agree with the idea that this could be an effective means of capturing and transferring tacit organizational knowledge. Identify people in the organization willing to share how they learned from others about how to do their jobs. Metaphors are a way to confront difficult organizational issues. Stories can only transfer knowledge if the listener is interested in learning from them.

Where can storytelling be effective? Igniting action in knowledge-era organizations Bridging the knowing-doing gap Capturing tacit knowledge To embody and transfer knowledge To foster innovation Enhancing technology e.g e-mail can connect people 24/7 Individual growth – avoids competition, mutual benefit Launching/Nurturing communities of practice thematic groups (World Bank) learning communities or learning networks (HP) best practice teams (Chevron) family groups (Xerox) Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Techniques for Organizing and Using Stories in the Organization Anthropological observation naïve interviewers asked innocent and unexpected questions caused the subjects to naturally volunteer their anecdotes (or story) or subjective views curiosity resulted in a higher level of knowledge elicitation Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Techniques for Organizing and Using Stories in the Organization Story-telling circles formed by groups having a certain degree of coherence and identity Methods for eliciting anecdotes: Dit spinning (fish tales or legend) - —represents human tendencies to escalate or better the stories shared previously Alternative histories - are fictional anecdotes which could have different turning points, for example on a particular project’s outcome. Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Techniques Shifting character or context - are fictional anecdotes where the characters may be shifted to study the new perspective of the story. Indirect stories -- allow disclosing the story with respect to fictional characters, so that any character similarities with the real-life character are considered to be pure coincidence. Metaphor -- provides a common reference for the group to a commonly known story, cartoon, or movie

Knowledge Representation through the use of Concept Maps based on interview sessions between a knowledge engineer and the domain expert, with the goal of jointly constructing an expertise model. Although computers may understand the resulting expertise models, these models may not directly meet the objective of capturing and preserving the expert’s knowledge so it can be transferred to others, or in other words, so others can learn from it. Based on Ausubel’s learning psychology theory Concepts, enclosed in circles or boxes. are perceived regularities in events or objects designated by a label Two concepts connected by a linking word to form a proposition, semantic unit or unit of meaning Vertical axis expresses a hierarchical framework for organizing the concepts inclusive concepts are found at the top, progressively more specific, less inclusive concepts arranged below relationships between propositions in different domains are cross-links Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Additional methods Interviews Video recording Observation Codification – writing down what one knows Group discussion Collaborative software tools such as wikipedia Etc

Concept Map about Concept Maps Concept maps as a knowledge-modeling tool. Concept maps, developed by Novak (Novak, 1998) aim to represent knowledge through concepts enclosed in circles or boxes of some types, which are related via connecting lines or propositions. Concepts are perceived regularities in events or objects that are designated by a label Simple concept map contains just two concepts connected by a linking word to form a single proposition, also called a semantic unit or unit of meaning. “Concept maps represent organized knowledge.”

Concept Map about Concept Maps Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

A Concept Map Segment from Nuclear Cardiology Domain Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Construction of Concept Maps Hierarchical the vertical axis expresses a hierarchical framework for organizing the concepts. More general, inclusive concepts are found at the top of the map with progressively more specific, less inclusive concepts arranged below them. Emphasize the most general concepts by linking them to supporting ideas with propositions. E.g Animal-> domestic -> cow

Semantic networks, also called associative networks, are typically represented as a directed graph connecting the nodes (representing concepts) to show a relationship or association between them. useful in describing, for example, traffic flow in that the connections between concepts indicate direction, but directed graphs do not connect concepts through propositions. Does not indicate generalization from specific to general

Knowledge Capture Systems: CmapTools To capture and formalize knowledge resulting in context rich knowledge representation models to be viewed and shared through the Internet Alleviates navigation problem with concept maps Serve as the browsing interface to a domain of knowledge Icons below the concept nodes provide access to auxiliary information Linked media resources and concept maps can be located anywhere on the Internet Browser provides a window showing the hierarchical ordering of maps

Segment from Nuclear Cardiology using CmapTools Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Explanation Subsystem using CmapTools Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Knowledge representation through context-based reasoning Tactical knowledge human ability that enables domain experts to assess the situation at hand (therefore short-term) myriad of inputs, select a plan that best fits current situation, and executing plan recognize and treat only the salient features of the situation gain a small, but important portion of the available inputs for general knowledge

Knowledge representation through CxBR Context - set of actions and procedures that properly address the current situation As mission evolves, transition to other context may be required to address the new situation For example knowledge sharing between supervisor and subordinate and between subordinate In hospital and academic institutions What is likely to happen in a context is limited by the context itself Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Knowledge representation through CxBR (context based reasoning) Mission Context - defines the scope of the mission, its goals, the plan, and the constraints imposed -- e.g drive to work (high way and local road) Main Context - contains functions, rules and a list of compatible subsequent main Contexts (e.g high way driving) Sub-Contexts - abstractions of functions performed by the Main Context which may be too complex for one function (e.g traffic light action) Only one action (e.g red light) occur but changed by events from the environment. The agent knows the events

Knowledge representation through CxBR Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Knowledge Capture Systems based on CxBR Context-based Intelligent Tactical Knowledge Acquisition (CITKA) uses its own knowledge base to compose a set of intelligent queries to elicit the tactical knowledge of the expert composes questions and presents them to the expert result is a nearly complete context base can be used to control someone performing the mission of interest in a typical environment Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Knowledge Capture Systems based on CxBR CITKA consists of four modules of independent subsystems: Knowledge engineering database back-end (KEDB) Knowledge engineering interface (KEI) Query rule-base back-end (QRB) Subject matter expert interface (SMEI) Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Barriers to the use of knowledge capture systems Barriers to the deployment of knowledge capture systems from two perspectives: the knowledge engineer who seeks to build such systems the subject matter expert, who would interact with an automated knowledge capture system to preserve his knowledge Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Barriers to the use of knowledge capture systems Knowledge Engineer requires developing some idea of the nature and structure of the knowledge very early in the process must attempt to become versed in the subject matter, or the nature of knowledge An automated system for knowledge capture, without a-priori knowledge of the nature, is essentially not possible Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Barriers to the use of knowledge capture systems From the point-of-view of the expert: need to take the initiative of learning how to interact with the system some people may be resistant to trying new things can be overcome, with adequate training and the utilization of user-friendly interfaces Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall

Using learning by observation capture knowledge Research on how humans and animals learn through observation has promise for development of knowledge based systems Use of learning through observation to automate the knowledge acquisition task Learning by observation shows promise as a technique for automatic capture of expert’s knowledge, and enable computers to automatically “learn”

Conclusions In this chapter we: Described knowledge capture systems design considerations specific types of such systems Discussed different methodologies and intelligent technologies used to capture knowledge concept maps as a knowledge-modeling tool context-based reasoning to simulate human behavior Explained how stories are used in organizational settings to support knowledge capture Becerra-Fernandez, et al. -- Knowledge Management 1/e -- © 2004 Prentice Hall