Transcript Lecture22

Trustworthy Semantic Web
Knowledge Management + E-Business +
Semantic Web = Semantic E-Business
Dr. Bhavani Thuraisingham
March 2010
3/31/2016 16:03
2
Outline of the Unit
0 What is Knowledge Management?
0 Basic concepts: Components and Models
0 Organizational Learning Process
0 Knowledge Management Architecture
0 Secure Knowledge Management and Trust Negotiation
0 Knowledge Models
0 Some efforts
0 Integration of KM with E-Business and Semantic Web
0 Reference: IEEE Transactions on Systems, Man and
Cybernetics, May 2006
0 Chapters 17 and 19 of the textbook
3/31/2016 16:03
3
What is Knowledge Management
0 Knowledge management, or KM, is the process through which
organizations generate value from their intellectual property and
knowledge-based assets
0
KM involves the creation, dissemination, and utilization of
knowledge
0 Reference: http://www.commerce-database.com/knowledge-
management.htm?source=google
3/31/2016 16:03
4
Knowledge Management Components
Knowledge
Components of
Management:
Components,
Cycle and
Technologies
Components:
Strategies
Processes
Metrics
Cycle:
Knowledge, Creation
Sharing, Measurement
And Improvement
Technologies:
Expert systems
Collaboration
Training
Web
3/31/2016 16:03
5
Knowledge Models
0 Level 1: Highest Level
- Mental models utilized by psychologists
- Social models (e.g. social network models) used by sociologists
0 Level 2: Mid-level
- Models utilized by expert systems
- Process modeling
0 Level: Bottom level
- Models understood by machines
- E.g., rule-based, frame-based, etc.
3/31/2016 16:03
6
Organizational Learning Process
Diffusion Tacit, Explicit
Identification
Creation
Source:
Reinhardt and Pawlowsky
Metrics
Integration
Modification
Action
also see: Tools in Organizational Learning
http://duplox.wz-berlin.de/oldb/forslin.html
3/31/2016 16:03
7
Six Principals of Effective Learning
• Effective Learning Requires:
 Understanding
1) Mental models, paradigms, context, observation,
assumptions, opinion, fact, truth
2) Systems Thinking - Variation
 Skills
3) Ability to challenge assumptions
4) Listen to Understand
 Process
5) Complete observe, assess (reflection, gain
understanding), design (develop theory, prediction, vision),
implement (test), cycle
6) Teach others
3/31/2016 16:03
8
Knowledge Management Metrics The Goal of Metrics
 Measuring Success (How am I doing?)
 Benchmarking (How am I comparatively doing?)
 Tracking Improvement (Am I getting better?)
 Direct
 future investment (technology, employees)
 strategy
 alignment (culture, incentives)
“One way to ensure your doing worse is to not measure” Adapted from Pressman
3/31/2016 16:03
9
Learning By-Product Measures
 Papers in Competitive Journals and Magazines
 Percentage New Technology compared to all Technology
 Process Cycle Time
 Employee Surveys
 Involvement with decisions
 Recognition for work achieved
 Access to information
 Rewarding risk taking
 Overall Satisfaction
 Employee Retention
 ‘Employee Suggestion Process
3/31/2016 16:03
10
Knowledge Management: Incentive-based
Approaches
Receiver
Teacher
Positive Incentives
Positive Incentives
• Knowledge Gained
• Can teach others
what is learned
Negative Incentives
• “Knowledge Transfer
Champion” prestige
•Can improve knowledge
Negative Incentives
•Time
•Time
•Unqualified teacher
•Students not willing to
learn
3/31/2016 16:03
Knowledge Management: Strategies, Processes,
Metrics and Tools
Knowledge Management: Within and Across
Corporations and Agencies
Strategies
e.g., Management
Plans; Policies;
Data sharing vs. Privacy
Metrics
e.g., web usage
Processes
e.g., best
practices
Tools
e.g., Semantic
Web
11
3/31/2016 16:03
12
Knowledge Management Architecture
Knowledge
Creation and
Acquisition
Manager
Knowledge
Dissemination
and Sharing
Manager
Knowledge
Representation
Manager
Knowledge
Manipulation
Manager
3/31/2016 16:03
13
Knowledge Exchange & Annotation
Engine (KEAN)
 Resides on any web-accessible knowledge base (any
intranet, www)
 Increases incentive to share information
 Author gets positive and negative feedback about
information that is submitted
 Feedback system - no more publishing documents that
disappear into the ether
 Prestige - top rated document views
 Quality filters steer user towards best information
 Domain specific instances of KEAN are created
 Works with Java enabled browser
3/31/2016 16:03
14
The Three Versions of KEAN Architecture
 Version 1 “beta version”
 No reuse
 Two-tiered
 Stored procedures
 Version 2 “newest version”
 GUI reuse via JavaBeans
 Two-tiered
 JDBC access to database
 Version 3 “final version”
 Logic reuse via Enterprise JavaBeans (EJB)
 Three-tiered
 CORBA access to objects
3/31/2016 16:03
15
Structure of Version 1 - Beta Version of KEAN
Stored
Procedures
(PL/SQL)
ORACLE WEB
SERVER
•Html
•JavaScript
•Applets
•ORACLE
“Thin http” client
Database tier
3/31/2016 16:03
16
Structure of Version 2
Code reuse with a two tier architecture
JDBC
ORACLE WEB
SERVER
•KeanBeans
•ORACLE
“Thick” client
Database tier
3/31/2016 16:03
17
Structure of Version 3
web Server
•KeanBeans
CORBA ORB
•EJB
“thin” client
middle tier
RDBMSs
and
OODBMS
server tier
3/31/2016 16:03
18
Secure Knowledge Management
0 Protecting the intellectual property of an organization
0 Access control including role-based access control
0 Security for process/activity management and workflow
- Users must have certain credentials to carry out an activity
0 Composing multiple security policies across organizations
0 Security for knowledge management strategies and processes
0 Risk management and economic tradeoffs
0 Digital rights management and trust negotiation
3/31/2016 16:03
19
Trust Management and Negotiation
0 Design a Trust Model
- Investigate the current trust models. Identify the inadequacies of current
trust models and design a model for the semantic web/DIVO
= Components include trust management, trust negotiation as well as
economic tradeoffs
0 Design a Language for specifying Trust policies
- Start with XML, RDF and Web Rules language and incorporate features
for trust management and negotiation
0 Design and develop techniques for enforcing the trust policies
- Automated Trust Negotiation: A attempts to access database D based on
access control policies; However before A can access D, triggers go off
and owner of D exchanges credential information with A (
3/31/2016 16:03
20
Knowledge Management for Coalitions
Knowledge for Coalition
Export
Knowledge
Export
Knowledge
Export
Knowledge
Component
Knowledge for
Agency A
Component
Knowledge for
Agency C
Component
Knowledge for
Agency B
3/31/2016 16:03
21
Status and Directions
0 Knowledge management has exploded due to the web
0 Knowledge Management has different dimensions
- Technology, Business
0 Tools are emerging
0 Need effective partnerships between business leaders, technologists
and policy makers
0 Major direction is integrating E-Business processes and semantic
web technologies for knowledge management
3/31/2016 16:03
22
Semantic E-Business
0 E-Business processes (e.g., order management, supply chain
management, contracts management, workflow management)
0 Service oriented architectures
0 Apply semantic web technologies such as XML, RDF, Ontologies
and RulesML to represent data and reason about the data for the ebusiness processes
0 Results in effective knowledge management as organization is
getting benefits
0 The topic is called Semantic E-Business
0 IEEE Transactions on Systems, Man and Cybernetics, March 2006
3/31/2016 16:03
23
Some Efforts - 1
0 Messaging (ebMS) : This is a specialization of web services
0
0
0
0
for business to business applications.
Business Process and Collaboration (ebBP) : This set of
specification enables collaboration among business partners.
Collaboration Protocol Profile and Agreement (CPPA) : Their
effort provides definitions for the sets of information used in
business collaborations.
Registry and Repository : The goal of this effort is to come up
with specification hat enable interoperable registries and
repositories
Core Components (CCTS) : This effort focuses on
technologies such as context and content assembly.
3/31/2016 16:03
24
Some Efforts - 2
0
Semantic web technologies have many applications in knowledge management. For
example, we need ontologies to capture the represent knowledge and reason about the
knowledge.
0
Paul Warren gives an example on how ”a political scientist, Sally who wants to research
the extent to which British Prime Minister Tony Blair's stance on Zimbabwe has changed
over a year and what factors might have caused that change.”
0
He further states that “in the world of the Semantic Web, Sally could search for
everything written by Blair on this topic over a specific time period. She could also
search for transcripts of his speeches. Information markup wouldn't stop at the article or
report level but would also exist at the article section level. So, Sally could also locate
articles written by political commentators that contain transcripts of Blair's speeches”
0
Now knowledge management also has applications for building the semantic web. For
example, prior knowledge captured as a result of knowledge management can be used
by agents to better understand the web pages. With respect to security, in the example
by Warren, confidentiality, privacy and trust policies will determine the extent to which
Sally trusts the articles and has access to the articles in putting together her report on
Tony Blair’s speeches.
3/31/2016 16:03
25
Some Efforts: OBELIX
0 Ontologies have also been developed for e-commerce applications specified
in languages such as RDF, RDF-S, OWL and OWL-S
0 For example, in the Obelix project a very good description of e-business and
ontologies is provided. The authors state that a problem with e-commerce is
the vague ideas that lack precise description they then discuses their
approach which they call e3value which is based on requirements
engineering and they define ontologies for e-commerce.
0 It is stated that “OBELIX is the first ontology-based e-business system of its
kind in the world to provide smart, scaleable integration and interoperability
capabilities”.
0 It is also stated that this project “ incorporates ontology management and
configuration, an e-business application server and ontology-based eapplication tools as well as an e-business library.”
0 OBLEIX is a European Commission project and the goal is to automate e-
business services in a semantic web environment which has come to be
called semantic e-business.