FB5807 4

1
FB5807 4

• IS Tools for Decision Making

DSS, ESS, Data Mining, Knowledge Management
Some slides adapted from Chris Wagner, 2005
2
Learning Objectives

• Understand systems and tools that can be used to
  support decision making by managers, groups or
  executives
• Understand executive roles from an information
  systems perspective.
• Recognise the importance of knowledge management
  functions

3
Motivation
Command-and-control system. Unknown situations
lead to delegation upwards. As the number of
unprecedented situations increases, information
flows increase, and speed of decision making
decreases.
4
As Uncertainty Increases

• Slow-down of decision making.
• Need for more distributed decision making.
• Need to authorize (empower) employees.
• Need to provide employees with necessary
  knowledge to make decisions.
• DSS, ESS and KMS all contribute to a solution
  here.

5
Decision Support Systems

• DSS serve the management level of an
  organisation, helping managers make decisions
  that are semi-structured, unique or rapidly
  changing and not easily specified in advance.
• DSS use internal and external information about
  competitors for instance.

6
DSS Components

• Database Data Warehouse
• current and historical data
• Online software tools for data analysis
  (including data mining tools)
• mathematical analytical models (e.g.
  statistical models) - building relationships
  between variables in the data
• User interface
• GUI simple intuitive for dumb executives

7
Does your Organisation use DSS? How?
8
Executive Support Systems (ESS)

• Not just information (EIS), but analytical
  processing too.
• Exception Reporting Trend Analysis
• Drill down
• Exactly what happened here? Who did what?
• Considering the last 4 years of data, in 400
  outlets worldwide
• Which are the most consistently slow-selling
  items?
• Which items get stolen frequently?
• What do Chinese shoppers buy?
• Did doubling the price of Chateau Robert 2001
  (Bordeaux) increase sales? Cross-market analyses.

9
Why Use an ESS?

• Provide easier, faster access to info
• Improve exec. efficiency and effectiveness
• Monitor organizational performance
• Improve communication
• Extract and integrate incompatible data
• Enrich execs mental model of their environment
• Competitive information
• Monitor external environment
• Support broadened span of control

10
Do you have access to Executive Support Systems?
What do you do with them?
11
Data Mining

• Data mining is named because of apparent
  similarities with the mining of a mountain for a
  vein of precious ore e.g. diamonds, rubies,
  etc.
• In both cases, you need to sift through an
  enormous amount of material (or else search it
  intelligently) to find the gems the value.

12
Data Mining Issues 1

• Data is often buried deep in space and time
• May be in a data warehouse or somewhere out
  there on the Internet
• Data mining tools are used to extract the
  valuable ores
• GeneMiner is one such tool, available from
  www.kDiscovery.com
• The miner is usually an end-user, often a manager
  or executive, not a programmer

13
Data Mining Issues 2

• Data mining tools can be combined with
  spreadsheets so as to enable easy analysis
• Unexpected and valuable results are a major
  benefit
• Serendipity
• Multiple computer processors may be needed, given
  the large amount of data
• Data Mining is often linked to a DSS/ESS.

14
Data Mining Techniques

• Case-based reasoning
• comparison against historical cases
• Neural computing
• pattern recognition and extrapolation, esp in
  financial services
• Intelligent agents
• remote access by agents of Internet databases

15
Data Mining Applications 1

• Retailing and sales prediction, inventory
  management, distribution schedules
• Banking forecasting of problems in loans and
  credit card fraud which customers best respond
  to new loan offers
• Maunfacturing predicting machine failures and
  identifying key factors that control optimisation
  of manufacturing

16
Data Mining Applications 2

• Brokerage/securities trading predicting price
  changes in bonds, stock fluctuations, when to
  buy/sell based on historical patterns
• Insurance Forecasting claim amounts and medical
  coverage costs predicting which people will buy
  new policies
• Police tracking crime patterns, locations,
  behaviours

17
Data Mining Applications 3

• Airlines where do our customers fly to? If they
  change airline, where is the final destination?
• Analysis of retail sales data to find seemingly
  unrelated products that are bought at the same
  time, e.g. baby diapers and beer.
• Planning of retail outlets as a result of pattern
  discovery beer and diapers should be next to
  each other.

18
What data mining applications can you see in your
organisations?
19
Knowledge Management

• Example 1
• I have been searching for a solution to a problem
  for 6 months. Eventually, I find the answer in an
  academic journal and the author is my
  colleague from two-doors away down the corridor!

20
Knowledge Management

• Example Two
• I am the senior partner of a global headhunting
  firm. We have reasonable information management
  (industry analysis, market research, resumé
  databases) but do nothing to tap into the vast
  knowledge resources held in the brains of our
  consultants. Each time a consultant leaves, our
  firms collective brain is drained. What can I do
  to manage our knowledge resources better?

21
Does your organisation manage knowledge?How?
22
KM Fundamentals ??

• What is knowledge ?
• Why is it important?
• What do we have to do in order to understand our
  knowledge resources
• How are we going to be able to manage our
  distributed knowledge resources?

23
KM

• Knowledge is information that is contextual,
  relevant, and recontextualisable for action.
• Knowledge is based on someones experience.
• Possessing knowledge gives one the opportunity to
  act.
• The financial value of knowledge is sometimes
  expressed as intellectual capital.
• Tacit or explicit?
• (Nonaka Takeuchi)

24
How do we conceptualise knowledge?

• As a formal organisational resource?
• As a community resource?
• As an individual resource?
• As something that can be codified in documents?
• As something that is best explained
  person-to-person?

25
Knowledge Needs, Validity Use

• What are the knowledge needs of employees?
• How quickly does knowledge change, degrade
• When is its use by date?
• What is its half life?
• What does knowledge recontextualisation involve
  and cost?

26
Knowledge Organisation and Delivery

• Hierarchies
• Communities
• Markets
• Codification-based systems
• Personalisation-based systems

27
Knowledge Hierarchies

• Specific knowledge that is customised for target
  users and often reused
• Hierarchies imply a consistent storage mechanism
  that is easily searchable
• High creation costs
• Accuracy, completeness and integrity (of
  knowledge and source) are important
• Quality is high, but validity is often short

28
Knowledge Communities

• Knowledge is shared among community members, with
  trust-supported sharing
• Community norms are influential
• A coordinator will facilitate the communitys
  access to knowledge
• Feedback mechanisms will validate the knowledge
• Quality is variable, and validity is often longer
• Short-validity knowledge requires too much effort
  to update on a regular basis.

29
Knowledge Markets

• A market will focus on capture, not creation of
  knowledge
• Each individual employee acts alone
• With little formal KM, there is little
  validation/organisation
• This reduces creation costs, but increases search
  and recontextualisation costs
• This is a chaotic bazaar, where quality is an
  unknown factor
• Buyer beware answers.google.com

30
Codification-Based Delivery 1

• Expert System
• Formally codified knowledge automated
  search/dissemination
• Knowledge Repository
• Text database of documents quite easy to locate
  knowledge
• Document Repository
• Text database of documents, but no specific
  knowledge examples

31
Codification-Based Delivery 2

• Exemplars and Templates
• Text database of best practices for specific
  tasks
• Exemplars are examples that illustrate best
  practices
• Templates involve step-by-step scripts
• Tips, Stories, Opinions, Principles, Heuristics,
  Patterns
• Example/scenario-based text similar to
  exemplars/templates, but less structured

32
Personalisation-Based Delivery

• Expert Directory
• Managed and validated database of people formally
  recognised as being experts
• People Directory
• Organised list of people with espoused interest
  in a particular area, but little
  validation/verification of their knowledge

33
Personalisation-Based Delivery

• Reference centre
• Chauffeured access point to knowledge
• Knowledge comes from a designated expert
• QA Forum
• Web-based discussion site or blog
• Community Calendar
• Shared calendar of events of interest to the
  community

34
KM, KS and Communities

• KM needs sharing of ideas.
• A notoriously difficult barrier to effective KM
  implementation.
• KS needs to be rewarded KS failure must also be
  rewarded.
• KS needs to be easy, not time/energy consuming.
• Sharing is often easier in communities.
• Do some communities find it easier to share than
  others?

35
Systematic Knowledge Processes

• Does the firm have systematic processes for
• Capturing, organizing and sharing
• external and internal knowledge?
• Are there processes for enhancing knowledge
  creation and innovation?
• Are there procedures governing the protection of
  knowledge assets?
• Does senior management actively promote a
  knowledge sharing culture?
• Are knowledge contributions measured or linked to
  financial performance indicators?

36
Business Knowledge Internal

• Management Technical Information for
  Decision-making
• Management information, in-house research,
  technical and product materials
• Rules Guidance for Operations and Management
• Description of objectives, process work flow,
  practice guidance (process description goals,
  quantitative and qualitative requirements, timing
  requirement, important matters), document
  templates and examples, authorizations and
  controls, rewards and punishments
• Management Experiences and Intelligence
• Best practices, case facts, expert channels

37
Business Knowledge External

• Macro-economy, Industries and Markets
• customers, competitors, technologies, products,
  market intelligence
• Commercial Knowledge
• finance, trading, investments, accounting, taxes,
  certifications, trade marks and patents, legal
  services, business consulting, public
  relationship, exhibitions, environmental
  protection, advertising, design and printing,
  packaging, software, asset trading and
  dispositions
• Business Operations and Management
• theories and methodologies, best practices, case
  facts (successes / failures)

38
Business Knowledge Major Types

• Working Guidance and Experiences, Document
  Templates and Examples, Document Records
• Laws and Regulations, Business Intelligence, News
  and Information
• Management Data, Business Analysis and Reports
• Theories and Methodologies, Practice Approaches
  and Case Facts
• Expert Channels and Knowledge Channels

39
People, Knowledge Technology

• In order to execute a knowledge based strategy,
  we need to think how to nurture people with
  knowledge
• Knowledge is most effectively applied through
  networks of people who collaborate with one
  another not through networks of technology
• KM is a primarily human-human process, supported
  by technology. Treating KM as a technical problem
  and finding a technical solution is likely to
  result in failure.
• KM strategies are more likely to be successful if
  they are driven by human needs for help in
  solving problems, not by knowledge being pushed
  at people.

40
KM and Reward Structures

• Both creators and users of knowledge should be
  rewarded.
• Mistakes are also a source of knowledge so
  reward their reporting
• Knowledge sharing should be recognised
  financially and publically
• Failure to use/share knowledge should be
  penalised
• Rewards can be designed at both individual and
  team levels
• Time must be allocated to knowledge creation and
  sharing.

41
KM, Top Management and Hiring Policies

• If people oppose KM blindly and try to destroy
  knowledge management efforts underway in the
  organisation, then dont promote or encourage
  them!
• Dont let KM initiatives be held back by old
  culture and old thinking
• All employees from the CEO downwards need to
  abandon the old and adopt the new
  enthusiastically if KM is to be successful

42
KM Initiatives and Alignment
43
Knowledge Management Framework
Leadership Strategy
Starting points
Adapt
Use
Create
Culture
Measurement
Value Proposition
Share
Identify
Collect
Organize
Technology
Based on a model co-developed by APQC AA, 1995
44
Same Purpose Different Paths
Employees write up consulting reports
KM team locates all such valuable sources
Documents are stored on a corporate portal
Indices, categoriz-ation, context is added
Users are provided with intelligent search
Knowledge portal is modified as use changes
Employees share advice via discussion boards
KM team locates all such discussions
Advice is categorized, reformatted
Web links to discussion board and categories
Old discussions are archived, while repeated
questions are transformed into FAQs
Collection of e-mail enquiries
KM program searches for patterns (text mining)
FAQs and answers stored in knowledge base
Knowledge base either makes suggestions about
best answers, or automatically answers 80 of
e-mail inquiries.
Knowledge is updated based on new inquiries
Collection of numerical transaction data
KM team searches for patterns (data mining)
Patterns are stored in rule form (knowledge base)
Patterns are reported as business rules, or can
be used to intelligently search through databases
(profiling)
Knowledge is updated based on new records
45
Other Resources

• Knowledge Maps
• Conversational Knowledge
• Google
• Now search results depend on networks of links
  between pages and page currency, not just keyword
  counts anymore.

46
(No Transcript)
47
Conversational Knowledge

• Blogs, Wikis, Email, Skype,
• Conversations become persistent,
  google-searchable, part of your knowledge network

48
Discussion Bobs Story

• What is his working environment?
• What are his knowledge tools?
• What benefits can the company achieve through
  people such as Bob?
• What are Bobs incentives to share knowledge?
  (and his sources)
• Would this work in Hong Kong?

49
Conversational Knowledge 2

• Id like all of you to download Skype (if you
  dont have it already) and then to have a Skype
  conversation with 1-2 other LPs.
• Try both the audio and textual tools
• And then cut/paste the text conversation into
  your blog!