(return to home lab)

 

 

PIERRE DERRIER - #203293

UNIVERSITAD DE LAS AMERICAS

Cholula, Puebla

MEXICO

 

 

 

 

 

 

 

KNOWLEDGE MANAGEMENT

READING REPORT :

 

An Intelligent Agent-Based Framework for Knowledge Management on The Web:
An Exploratory Study of A Virtual Team in Designing A Multimedia System

Seung Ik Baek, Jay Liebowitz, and Srinivas Y. Prasad
Mary J. Granger
Management Science Department
School of Business and Public Management
The George Washington University
Washington, DC 20052

CONTEXT DESCRIPTION

The recent proliferation of personal computers and communication networks has enabled organizations to acquire and retain their distributed organizational structures. Using a computer network, geographically distributed people with common goals can communicate, coordinate, and collaborate their work efforts across time and space barriers. These "virtual teams" can bring together the right mix of people who have the appropriate set of knowledge, skills, information, and authority to solve difficult problems quickly and easily.

The challenge that modern organizations face is to turn the scattered, diverse knowledge of their knowledge workers who are working in a virtual team into a well-structured knowledge repository.

 Knowledge Management (KM) is suggested as a methodology for creating, maintaining and exploiting a knowledge repository. KM is defined as the collection of processes that support the creation, dissemination, and utilization of knowledge between appropriate individuals, groups within an organization, and independent organizations

 

ABSTRACT OF THE PAPER

This research focus on developing a conceptual model for KM and a framework for the roles of intelligent agents in a conceptual KM model. Furthermore, it will implement and evaluate the intelligent agent-based framework on the Web under a collaborative environment for designing a multimedia system.

 

SUMMARY OF THE PAPER

 

A. CONCEPTUAL MODEL

Spek and Spijkervet (1996) identify three basic activities necessary to build a well-structured knowledge repository:

1. Creating knowledge
2. Securing/combining knowledge
3. Distributing/retrieving knowledge.

 

1. Creating Knowledge

            Knowledge is created by designers, developers and users of the system. The knowledge is refined continually by other group members throughout a design process. A highly expressive, precisely defined set of attributes for design knowledge representation can facilitate the ability to create new knowledge in a knowledge repository.

2. Securing/Combining Knowledge

All new knowledge must be validated by all team members, in order to make sure they understand it clearly, and can freely integrate their own knowledge into the new knowledge. All these can happen by storing and indexing knowledge properly.

3. Distributing/Retrieving Knowledge

The new knowledge must then be shared and easily accessible. In a rapidly changing environment, keyword-based knowledge retrieving is no longer productive. The dynamic nature of such an environment requires content-based, context-based knowledge retrieving.

 

B. IMPLEMENTATION WITH KM AGENTS

To enhance knowledge flow in a conceptual KM model, three KM agents are designed and implemented:

1.      User agent

2.      Knowledge manager

3.      Knowledge agent.

KM Activities

KM Agents

Major Functions

Creating Knowledge

User Agent

Remember all KM activities of users; Dynamically organize a person's agenda.

Securing Knowledge

Knowledge Agent

Index knowledge; Detect inconsistency; Save, retrieve, and update knowledge from a knowledge repository.

Distributing Knowledge

Knowledge Manager

Monitor all changes that occurred in a knowledge repository; Provide critics to new knowledge 

Retrieving Knowledge

Knowledge Manager

Reformulate queries based on an ontology; Determine the most favored alternative based on preference weighting and ranking.

 

C. VALIDATION STRATEGY

            This section focus on a way for validate the framework in a real-world setting. For that, it lists the KM activities of designers as they use an intelligent agent-based KM system.

For each, it proposes perfomance evaluation criterias :

KM Activities

Related Theories

Theoretical Propositions

Design Features

Performance Evaluation Criteria

Knowledge Creating

Gestalt Theory (Mayer, 1992)

In Gestalt Theory, problem representation rests at the heart of knowledge creating activity. The theory suggests that tools and techniques should be invented for helping people to represent problems in useful way. 

Storyboard-Based Knowledge Creating

Ease of Use of Storyboards; Clarity/Ambiguity of Storyboards; Richness of Storyboards

Knowledge Securing

IPS Model (Newell & Simon, 1972)

The IPS model assumes that human memory consists of two major components: long-term memory and short-term memory. 

Content-Based Indexing for Storyboards and Feedback Messages

Easy of Storyboard Editing; Effectiveness of Intelligent Access Control

Knowledge Distributing

Information Theory (Shannon, 1948)

Information Theory argues that the ability of individuals to generate and transmit knowledge has the potential to promote interdependency among individuals. It states that mutual awareness is an important issue for supporting collaboration.

Feedback

Timeliness/Responsiveness of Feedback; Negotiation Productivity

Knowledge Retrieving

Cognitive Flexibility Theory (Spiro et al., 1988)

Cognitive Flexibility Theory states that, if users can access various perspectives for solving a problem, they might get a deeper, clearer understanding about the problem. Because of the limited capacity of human memory, too much knowledge makes users experience cognitive overload problems.

Hypermedia-Based Knowledge Presentation

Cognitive Overload; Intuitive/Disorientation of Interface; Flexibility

 

REFERENCES

Full paper 
“An Intelligent Agent-Based Framework for Knowledge Managementon The Web: An Exploratory Study of A Virtual Team in Designing A Multimedia System”

 

 

Pierre Derrier    #203293

pierre.derrier@laposte.net

 

(return to home lab)