BCIS 401 Information Systems for Management

1
BCIS 401 Information Systems for Management

• Knowledge Management
• Dr. J. Affisco
• Fall 2001

2
Why Knowledge Management?

• Knowledge is the property of the individual
• Successful practices rarely transferable
• Knowledge is embedded and hard to extract
• Large amounts of time spent reinventing the
  wheel

3
Why Knowledge Management?

• Fortune 500 companies wasted 12 billion in 1999
  as employees duplicated one anothers work
• 90 of 800 North American and European Companies
  were working on some aspect of knowledge
  management
• Ford estimated that in 1997-99 knowledge
  management initiatives resulted in cost savings
  or additional revenues of 914 million

4
Why Knowledge Management?

• Chevron estimates it has saved more than 650
  million since 1991 by sharing best practices
  among mangers in charge of energy use at its oil
  refineries
• Texas Instruments reports savings of more than 1
  billion by disseminating best practices
  throughout its 13 semiconductor plants
• By late 1999 1/3 of top 1,000 largest U.S.
  companies had begun knowledge mgt. initiatives
  by 2003 more than half will have done so.

5
Types of Knowledge

• Explicit Knowledge
• Easily collected and organized
• Transferred through digital means
• Tacit Knowledge
• Personal context-specific
• Process knowledge
• Hard to formalize and communicate

6
Knowledge Management Strategies

• Sets forth criteria for choosing
• What knowledge a firm plans to pursue
• How firm will capture and share it
• Generally determined through strategic audit
• What sorts of knowledge are critical to support
  business positioning
• Who needs to have what information?
• When do they need to know it?

7
Knowledge Management Processes

• Generating
• Organizing
• Developing
• Distributing

8
Generating

• Identifying the desired content
• Getting people to contribute ideas
• On-line discussions
• Submitting deliverables that have emerged from
  other work

9
Generating

• Buy or Rent
• Research and Development
• Shared Problem Solving
• Adaptation
• Communities of Practice

10
Organizing

• Organizing collected data so it can be
  represented and retrieved electronically
• Knowledge sharing systems or tools
• Knowledge bases
• Navigational devices
• User interfaces
• Taxonomies

11
Organizing - Knowledge Bases

• Unfiltered
• Archive documents directly
• Many-to-many communication without intervention
  by others
• Filtered
• Content screened, distilled, and approved for
  use by recognized experts
• Material continually refreshed to maintain its
  currency

12
Developing

• Selection and further refinement of material to
  increase its value for users
• Subject matter experts review work done by others
  such as editors
• Results include
• Final content/form of expert material
• Knowledge objects

13
Distributing

• How people get access to material
• User friendliness
• Encouraging use and reuse of knowledge
• Types of systems
• Push - Sends large masses of information out to
  users
• Pull - Users call on the knowledge base to draw
  material out
• Targeted Push - Proactively deliver material that
  is context sensitive

14
Types of Decision Support

• Decision Support Systems
• Expert Systems/Knowledge Based DSS
• Group Decision Support Systems
• Executive Support Systems

15
Decision Support Systems
16
Phases of Decision Making Process

• Intelligence
• Design
• Choice

17
Intelligence

• Searching the environment for conditions calling
  for decisions
• Raw data are obtained, processed, examined for
  clues that may identify problems.

18
Design

• Inventing, developing, analyzing possible
  courses of action.
• This involves processes to
• Understand the problem
• Generate solutions
• Test solutions for feasibility

19
Choice

• Selecting a particular course of action from
  those available
• A choice is made and implemented

20
Levels of Org Decision Making
Information Characteristics Task
Variable Strat. Plan. Mgt. Control Operational
Control Accuracy Low High Level of
Detail Aggregate Detailed Time
Horizon Future Present Frequency of Use
Infrequent Frequent Source
External Internal Scope of Info.
Wide Narrow Type of
Info Qualitative
Quantitative Age of Info. Older

Current
21
A MIS/DSS Framework
22
Three DSS Levels
Specific DSS
DSS Generator
DSS Tools
23
Relating DSS Levels and Roles
Manager (user)
Specific DSS
Intermediary
DSS Generator
DSS Builder
Technical Supporter
Toolsmith
DSS Tools
24
Data Base
Model Base
DBMS MBMS
DSS
DGMS
Task
Environment
User
DSS Components
25
DSS Development Process
Systems Development Process
Builder and User agree on small significant sub-
problem.

• Analysis - What application is to
  be supported?

Design and develop an initial system to support
decision making required by sub-problem.

• Design - What is the best way to support
  application?

• Construction - Build the designed system.

Use the system for a short period of time.
Evaluate the system. Modify the system.
Incrementally expand the system.

• Implementation - Apply the system.

26
Expert Systems
27
Definition

• Expert Systems are special-purpose computer
  programs which use knowledge and reasoning to
  perform complex tasks in a specific problem
  domain at a level of performance usually
  associated with an expert in the domain.

28
Purpose

• Originally designed to replace domain experts.
• Today viewed as Knowledge-based Decision Support
  Systems
• System supports Managerial Decision Making with
  the capability to process knowledge in addition
  to quantitative data.

29
Components of ES

• Knowledge Base
• Inference Engine
• User Interface

30
Expert Systems Architecture
31
Knowledge Base

• Repository of domain-specific knowledge
• Knowledge needs to be represented and employed in
  a form that can be used for reasoning.
• Knowledge structures
• Facts
• Rules
• Frames

32
Facts

• General statements of truth that may be either
  temporary or permanent knowledge

33
Rules

• Knowledge structure of the form if-then
• The if statement represents a premise.
• The then statement represents a conclusion.
• As rules are processed, if the premise is true
  then the conclusion indicates some action to be
  taken.
• A rule is proved when the premise of the rule
  matches known facts.

34
Frames

• A way of packaging knowledge about one object.
• Are composed of slots in which data or
  characteristics associated with specific objects
  are stored.
• Frames are organized in a hierarchy which allows
  for sharing of knowledge through the property of
  heredity.

35
Rules vs Frames

• Frames are especially efficient for packaging
  knowledge and handling the storage and retrieval
  of that knowledge.
• Rules work best at making deductions.

36
Inference Engine

• Its task is to process the domain knowledge
  contained in the knowledge base to arrive at a
  solution to the problem.
• Inference engine combines facts and rules through
  an inference process to arrive at conclusions.
• Inference techniques
• Forward chaining
• Backward chaining

37
Forward Chaining

• Begins with known facts and the rule set and
  attempts to deduce new facts which may eventually
  lead to the deduction of the goal.
• Inference engine cycles through the rules until
  one is found whose premises matches a fact. This
  rule is then proved or fired, and the conclusion
  is added to the fact base.
• Process continues until the implication of the
  conclusions reached are sufficient to provide a
  solution.

38
Backward Chaining

• Inference processes work backwards from the goal.
• Inference takes the goal as a hypothesis and then
  seeks to prove a series of subgoals working
  backward from the goal.
• This is done recursively until all subgoals that
  are required for the goals existence are proven.

39
Group Decision Support Systems
40
GROUP DECISION-SUPPORT SYSTEM (GDSS)

• INTERACTIVE COMPUTER-BASED SYSTEM FACILITATES
  SOLUTION OF UNSTRUCTURED PROBLEMS BY DECISION
  MAKERS WORKING AS GROUP

41
TOOLS OF GDSS

• ELECTRONIC QUESTIONNAIRES
• ELECTRONIC BRAINSTORMING
• IDEA ORGANIZERS
• QUESTIONNAIRE TOOLS
• TOOLS FOR VOTING, SETTING PRIORITIES

42
TOOLS OF GDSS

• STAKEHOLDER IDENTIFICATION ANALYSIS TOOLS
• POLICY FORMATION TOOLS
• GROUP DICTIONARIES

43
ELECTRONIC MEETING SYSTEM (EMS)

• COLLABORATIVE GDSS USES INFORMATION TECHNOLOGY TO
  MAKE GROUP MEETINGS MORE PRODUCTIVE FACILITATES
  COMMUNICATION DECISION MAKING

44
HOW GDSS ENHANCED DECISION MAKING

• IMPROVED PRE-PLANNING
• INCREASED PARTICIPATION
• OPEN, COLLABORATIVE ATMOSPHERE
• IDEA GENERATION FREE OF CRITICISM
• EVALUATION OBJECTIVITY
• IDEA ORGANIZATION EVALUATION

45
HOW GDSS ENHANCED DECISION MAKING

• SETTING PRIORITIES DECISION MAKING
• DOCUMENTATION OF MEETINGS
• ACCESS TO EXTERNAL INFORMATION
• PRESERVATION OF ORGANIZATIONAL MEMORY

46
An Illustration of the Use of GDSS
47
The Problem

• Plant Location
• In practice, a cross-functional team of senior
  executives typically makes plant location
  decisions.
• In arriving at this decision the team looks at a
  number of conflicting variables and factors, both
  quantitative and qualitative.

48
The Objective

• To provide training for executives in plant
  location strategic decision making so as to
  improve their performance when faced with such a
  decision.

49
The Scenario

• A general scenario describing the industry,
  competitive environment, government and legal
  environment, and position of various other
  stakeholders is provided prior to the group
  support sessions.
• The exercise proceeds in two phases

50
Phase I

• The executive group is broken into three role
  playing groups
• Corporate management
• State and local government
• Environmental stakeholders
• A more specific scenario in terms of the firms
  products and processes is distributed with
  instructions to read it from the point of view of
  their role playing group.

51
Phase I - Individual GSS Sessions

• Electronic brainstorming is used to help the
  participants develop ideas about the plant
  location problem from their respective roles
  point of view.
• Idea organizer is used to help crystallize the
  ideas already generated into a set of critical
  issues for each of the three constituencies.
• Vote then is used by to prioritize the critical
  issues previously developed.

52
Phase I - Individual GSS Sessions

• The group is given information on three potential
  location sites and are then asked to evaluate
  the sites based on the critical issues they
  developed.
• The alternative evaluation tool is used to help
  rank order the competing location sites.
  However, the ranking is not revealed to the
  participants at this point.
• A transcript of the sessions is used to help
  debrief the groups on the decision making process
  as well as the decision content.

53
Phase II - Metagroup GSS Session

• The sets of critical issues identified, in the
  earlier sessions, by the groups representing the
  three constituencies are discussed.
• The stakeholder identification tool helps the
  group to analyze the impact of policies relating
  to the critical issues on the stakeholders.
• The policy formulation tool is used to facilitate
  the arrival at a consensus on which issues remain
  critical to the group.

54
Phase II - Metagroup GSS Session

• Vote may be used once again to arrive at a final
  prioritization.
• Having achieved this understanding of their
  position on the critical plant location factors,
  the metagroup is now given the information on the
  same three sites as during the individual group
  sessions.
• Using the appropriate tools, the metagroup is
  asked, for a second time, to rank order the
  alternatives. This time they are given the final
  ranking.

55
Phase II - Metagroup GSS Session

• In addition, at this time they are given the
  final ranking which resulted from each of the
  individual group decision sessions.
• A comparison of these rankings is used to fuel
  the debriefing discussion. As part of the
  debriefing we can compare the processes and
  results of the individual groups with those of
  the metagroup.
• A discussion of expectations versus reality is
  generally useful.