Transcript Future of Knowledge Management Systems

Future of Knowledge
Management Systems
Logan Buchanan
December 8, 2005
References
•
•
•
•
•
•
•
•
•
•
•
•
Ackerman, M. S (1994). Augmenting Organizational Memory: A Field Study of Answer Garden.
Ackerman, M. S. & McDonald, D.W. (1996) Answer Garden 2: Merging Organizational Memory with
Collaborative Help.
McDonald, David & Ackerman, Mark. (2000)Expertise Recommender: A Flexible Recommendation System
and Architecture. Proceedings of CSCW'00. ACM Press.
Bayardo, R., Agrawal, R., Gruhl, D., Somani, A. (2002) YouServ: A Web Hosting and Content Sharing Tool
for the Masses. WWW2002, May 7-11, 2002, Honolulu, Hawaii, USA
Malone, Lai, & Fry. Experiments with Oval: A Radically Tailorable Tool for Cooperative Work
Lieberman, H. (1997). Autonomous Interface Agents. ACM Conference on Human-Computer Interface
[CHI-97], Atlanta, ACM Press.
Voss, A. and Kreiflets, T. (1997) SOAP: Social Agents Providing People With Useful Information.
Proceedings of GROUP'97, ACM Press, pp. 291-298.
Shardanand, U. & Maes, P. (1995) Social information filtering: algorithms for automating "word of mouth".
Conference proceedings on Human factors in computing systems. ACM Press.
Maes, P. (1994) Agents That Reduce Work and Information Overload. Communications of the ACM, 37(7),
31-40.
Berners-Lee, T., Hendler, J. and Lassila, O. (2001) The Semantic Web. Scientific American, May 2001.
Stojanovic, Nenad & Handschuh, Siegfried. (2002) A Framework for Knowledge Management on the
Semantic Web. Proceedings of the World Wide Web conference 2002. Honolulu.
Yu, Bin & Singh, M. (2002) An Agent-Based Approach to Knowledge Management. CIKM'02. McLean, VA.
ACM Press.
Answer Garden
• Field study
 Group of users and group
of experts.
• Users
 Need for speed
 Social network and
information retrieval system
worked well together.
 Status implications.
• Experts
 Some refused to answer
based on their workload.
 Incentives questionable.
• Answer Garden 2
 No separate groups.
 Contextualize
answers.
 Escalation.
 Being anonymous is
optional.
Expertise Recommender
• Architecture open and flexible enough to
address different organizational environments.
• Organizationally specific implementations.
• User profiles created from organizationally
relevant data sources such as work products.
• Allows for user choice.
• Recommendations can be revisited and
escalated if necessary.
YouServ
• Pool existing desktop computing resources for
high quality web hosting and file sharing.
 Assigned a domain name.
 Pool resources of a group - content can be accessed
when your computer is off.
 Can publish even if behind a firewall.
 Low cost.
• Alternative to e-mailing attachments.
• Photo sharing is predicted most common use.
Oval
• Create applications by combining and modifying
objects, views, agents, and links (or Oval).
• Tailorable = end users can modify a working
system (such as a spreadsheet), changes are
made in the context of a working application
• Radical = large changes can be made
• User interface is simple and provides a “large
amount of functionality for creating and
modifying a wide range of applications”
Oval cont.
•
•
•
•
•
•
•
gIBIS - a tool for helping a group explore and capture the qualitative
factors that go into making decisions
Sibyl - goals for decisions and previous decisions
Coordinator - an e-mail based system that helps people structure
conversations and track tasks.
Lotus Notes
Intelligent Lens - automatically filters and sorts incoming e-mail
Answer Garden
And… a database of people, an organization chart, a project management
system, a system for tracking software bug reports, a system for
supporting marketing decision making, a system for supporting quality
management processes in manufacturing, and a workflow system for
purchase-order approval.
Autonomous Interface Agents
• Interface agents = software that actively assists a user in
operating an interactive interface
• Autonomous agents = software that takes action without
user intervention and operates concurrently
• Letizia
 An autonomous interface agent for Web browsing.
 Records URLs chosen by the user and reads the pages to
compile a profile of the user’s interests.
• Work best in situations where their decisions are not
critical, such as web browsing.
SOAP
• Social agents interact with one another on
behalf of their clients
 Examples: agents for electronic markets, workflow
agents, information brokers
• New users register in order to obtain a personal
user agent. A user agent links the user and the
other agents.
• For each query, a task agent is created.
• There are also group agents, recommender
agents, and search agents, directory agents, etc.
Social Information Filtering
• Social information filtering automates the process of
“word-of-mouth” recommendations.
 General trends and patterns within the preferences of a person
and between groups
 Considers thousands of people and thousand of items
• Personalized recommendations from databases based
on similarities between the interest profile of the user
and those of other users.
• Ringo, makes personalized recommendations for music
albums and artists.
• As more people use the system, Ringo is able to make
better predictions.
Agents
• Building Agents
 Two main problems
 Competence: how does an agent acquire knowledge?
 Trust: how can the user feel comfortable delegating tasks to
an agent?
 One approach (as used in Oval) is to make the enduser program the interface agent. This does not meet
the competence criterion.
 Another, the knowledge-based approach, gives the
agent extensive background knowledge about the
application and the user. Both competence and trust
are problems in this approach.
• Use machine learning techniques
The Semantic Web
• An extension of the current Web, in which
information is given well-defined meaning.
• Computers must have access to
structured collections of information.
• XML and RDF
• Ontologies
• Digital signatures
KM on the Semantic Web
• Semantic Web can serve as a platform for
developing knowledge management systems.
• Problem: How to represent knowledge in a
machine-understandable form, so that
appropriate knowledge can be found by agents?
• Use a conditional statement for the semantic
annotation of knowledge sources. Statements
used in the annotation can be put into the
context of each other, which leads to efficient
searching.
An Agent Based Approach to
Knowledge Management
• MARS, a multiagent referral system for
knowledge management
 MARS assigns an agent to each user
 Agents facilitate their users’ interactions
 Manage their personal social networks
 Agents cooperate with one another