Knowledge Management in Theory
and Practice
Lecture 8: KM Tools
Overview
KM Tools and Techniques
Knowledge mapping
Community of Practice mapping
Knowledge Taxonomies
Knowledge Repository Design
Groupware and collaboration tools
Knowledge repository tools (intranet, www, portal…)
EPSS (task support tools)
AI-based tools (e.g. CBR)
2
What does the end product look
like? Some examples
http://firefighternearmiss.com
3
Firefighters Nearmiss
Reporting System
4
Submitting a report
5
6
Age
Gender
Number years experience
Were you involved in the event? 7
Date, day of the week, approx. time?
How long for first crew members to arrive?
Key factors: severity, risk of life? Smoke? 8
9
10
11
Analytics
12
13
14
15
16
KM at the World Bank
17
World Bank KM Stories
18
Virtual Communities and "Low-Tech" Tools:
Lessons Learned at the World Bank Story
19
Enterprise KM Architecture
Data Layer
Unifying abstraction across different types of data with
potentially different storage mechanisms (database,
textual data, video, audio)
Process Layer
Describes the logic that links data with the use made by
people or other systems of that data
User Interface
Provides access for people to the information assets of
the enterprise via logic incorporated in the process layer
20
KM Enterprise Architecture
Profiles for Unifying User Interface User views
Personalization Or representations
UI Layer
Applications Functions for KM
Process
Help Locate Record Find Alert to
Layer
System Experts B P s Associations New Factors
Metadata
Data
Data Sources Data Types Data Formats Layer
21
Tools and Techniques for each
step of the KM cycle
Capture and store
Search and retrieve
Publish and disseminate
Structure and navigate
Use and apply
22
Checklist for KM Technologies
Capture and Store
Can define templates,
Can identify authors,
Can manage work steps to create, update, review
and approve end products
Shared capture processes (e.g. multiple authors)
Can capture and group units of knowledge in a
shareable repository
23
Checklist (con’t)
Search and Retrieve
Search and query without familiarity with content nor
location
Transparent access to disparate data sources
Efficient access through indices, data warehousing, data
marts or document repositories
Access to richer media and content such as text, video,
audio and images
Multiple search techniques such as simple Boolean, fuzzy,
conceptual or contextual, and feature detection searches
and complex natural language queries
24
Checklist (con’t)
Disseminate/publish
Routing and delivery of information to those
who have a need and the notification of
subscribers
Email, workflow, push technology to notify of
changes, of newly posted information, expired
subscriptions and expired materials
Pattern matching against user profiles (including
structured or adaptive profiles)
25
Repository Design
Transport
Rail Community of Practice
Transports
Canada Canada
ACTIONS Rail Industry
What’s New TC Headquarter TC Regions Links Reports Members Maps
Repository
Upcoming Events
Administration
Safety Related News
Simple
Search
One dead, 96 hurt as Amtrak train derails in Iowa
Advanced
Search Latest Accident Reports
Help New Publications
Glossary
New Members
26
Products – Repository Design
What’s New TC Headquarter TC Regions
Rail Industry
Links Reports Members Maps
27
Products – Repository Design
Transport
Rail Community of Practice
Transports
Canada Canada
ACTIONS TC Regions Rail Industry
What’s New TC Headquarter Links Reports Members Maps
Harmonization Studies (NAFTA)
Repository
Administration
Trend Analysis on Safety, Efficiency
Simple
Search
Research Reports
TRANSPORTATION - ENERGY AND ENVIRONMENT
Advanced
Search
Help
Glossary
28
TC Regions Rail Industry
What’s New TC Headquarter Links Reports Members Maps
29
Checklist (con’t)
Structure and Navigate
Provide a classification scheme for the organization’s knowledge
assets - hierarchies, taxonomies, semantic nets
Provide and a means to effectively navigate the structure using a
visual or textual UI path; enable multiple views
Index the explicit information content
Build electronic linkages from classification scheme to relevant
knowledge assets
Identify of human experts and their areas of expertise; id of users
and their CoPs
Link from UI topics to related KM content, people and processes
30
Taxonomies
Taxonomies are ways to structure vast amounts of information, e.g.
Dewey system, SIC, animal kingdom
Multiple parallel taxonomies can co-exist, e.g.
from a product point of view
from a process point of view
from a R&D point of view
The “first cut” at a taxonomy should be done by a domain expert
and KM expert
Taxonomies can be living entities (updateable)
Content that is mapped by the taxonomies can be automatically
“refreshed” and sorted
31
Example: epicurious.com
Course Dietary Ingredients Type of
Meal type Concerns cuisine
Dinner Vegan Chicken Italian
Lunch Low fat Fish Mexican
Breakfast Low salt Beef French
Snack Low sugar Turkey Thai
Appetizer Gluten free Vegetable Indian
Dessert… Kid friendly… Pasta… Mediterranean…
32
Checklist (con’t)
Share and Collaborate
Connect people with other people via groupware
Major KM technology – stimulates collaboration
Real-time application sharing and videoconferencing, chat rooms
Exchange of info as well as sharing creation of products, sharing
workspace
Ability to link to experts on-demand if they are online
33
Checklist (con’)
Synthesize
Discovery of new knowledge and insights from
available information: BI, data, skill, text
mining
Extracting data, downloading data for user
analysis and reuse, visual representation of
trends and patterns
34
Checklist (con’t)
Profile and Personalize
Align and group people into work and interest communities and
with information, objects or interests
Self-selected alignment, peer-selected alignment or enterprise-
determined alignment or automated alignment based on history of
usage
Filter incoming information to match user needs
Part of security program
Automated agents = s/w that acts as an intermediary for a person
by performing some activity. Agents can learn an individual’s
preferences to deliver knowledge to them “at point of need” – to
search, retrieve, synthesis, recommend on behalf of the user
35
Checklist (con’t)
Solve or Recommend
Encode knowledge in a model that produces a solution or
recommendation
E.g. credit-risk assessment, insurance underwriting,
equipment diagnosis
Rule-based systems, case-based reasoning systems,
neural networks used for this highly task-oriented type of
knowledge
Can be triggered by workflow rule and then delivered to
individuals or groups based on their profiles
36
Additional tools for the digital
workplace
Big data and analytics – AI, pattern recognition, data
mining
Visualization tools – knowledge maps are conceptual
representations of knowledge
Crowdsourcing – input from online community
Mobile technologies – increasing use of smartphone,
tablets for knowledge management
Collaboration tools (SharePoint)
Videos – e.g. for exit interviews
37
What about social media?
Wikis KM is inherently about
Twitter collaboration so it is a perfect fit
with social media
Blogs
KM also leads to greater
Web 2.0
transparency – again a good fit
FaceBook Last but not least – collaborative
LinkedIn content creation can only fuel KM
Others... implementation in both
organizational and in networks
38
Next:
KM Strategy
39
