AGENTS THAT HELP TO DETECT TRUSTWORTHY

KNOWLEDGE SOURCES IN KNOWLEDGE MANAGEMENT

SYSTEMS

Juan Pablo Soto, Aurora Vizcaíno, Javier Portillo-Rodríguez and Mario Piattini

Alarcos Research Group, Information Systems and Technologies Department, Indra-UCLM Research and Development

Institute, University of Castilla – La Mancha, Ciudad Real, Spain

Keywords: Knowledge Management Systems, Multi-agent architecture, Software Agents.

Abstract: Knowledge Management is a critical factor for companies worried about increasing their competitive

advantage. Because of this companies are acquiring knowledge management tools that help them manage

and reuse their knowledge. One of the mechanisms most commonly used with this goal is that of

Knowledge Management Systems (KMS). However, sometimes KMS are not very used by the employees,

who consider that the knowledge stored is not very valuable. In order to avoid it, in this paper we propose a

three-level multi-agent architecture based on the concept of communities of practice with the idea of

providing the most trustworthy knowledge to each person according to the reputation of the knowledge

source. Moreover a prototype that demostrates the feasibility of our ideas is described.

1 INTRODUCTION

Knowledge Management (KM) is an emerging

discipline considered a key part of the strategy to use

expertise to create a sustainable competitive

advantage in today’s business environment. Having

a healthy corporate culture is imperative for success

in KM. Zand (1997) claims that bureaucratic

cultures suffer from a lack of trust and a failure to

reward and promote cooperation and collaboration.

Without a trusting and properly motivated

workforce, knowledge is rarely shared or applied,

organizational cooperation and alignment are

nonexistent.

Certain systems have been designe

d to assist

organizations to manage their knowledge. These are

called Knowledge Management Systems (KMS).

KMS, described in (Alavi & Leidner, 2001), as an

IT-based system developed to support/enhance the

processes of knowledge creation, storage/retrieval,

transfer and application. An advantage of KMS is

that staff may also be informed about the location of

information. Sometimes the organization itself is not

aware of the location of the pockets of knowledge or

expertise (Nebus, 2001). Moreover, a KMS is able to

provide process improvements: it is better at serving

the clients, and provides better measurement and

accountability along with an automatic knowledge

management.

However, developing KMS is not a simple task

nce knowledge per se is intensively domain

dependant whereas KMS are often context specific

applications. KMS have received certain criticism as

they are often installing in the company thus

overloading employees with extra work, since

employees have to introduce information into the

KMS and worry about updating this information.

Moreover, the employees often do not have time to

introduce or search for knowledge or they do not

want to give away their own knowledge and or to

reuse someone else’s knowledge (Lawton, 2001). As

is claimed in (Desouza et al, 2006) “employees resist

being labeled as experts” and “they do not want their

expertise in a particular topic to stunt their

intellectual growth”. Because of this resistance

towards sharing knowledge, companies are using

incentives to encourage employees to contribute to

the knowledge growth of their companies (Huysman

& Wit, 2000). Some of these incentives are

organizational reward and allocate people to projects

not only to work but also to learn and to share

experiences. These strategies are sometimes useful.

However, they are not are a ‘silver bullet’ since an

employee may introduce information that is not very

useful with the only objective of trying to simulate

that s/he is collaborating with the system in order to

219

Pablo Soto J., Vizcaíno A., Portillo-Rodríguez J. and Piattini M. (2007).

AGENTS THAT HELP TO DETECT TRUSTWORTHY KNOWLEDGE SOURCES IN KNOWLEDGE MANAGEMENT SYSTEMS.

In Proceedings of the Second International Conference on Software and Data Technologies - PL/DPS/KE/WsMUSE, pages 219-226

DOI: 10.5220/0001327602190226

 SciTePress

generate points and benefits to get incentives or

rewards. Generally, when this happens, the

information stored is not very valuable and it will

probably never be used. Based on this idea we have

studied how the people obtain and increase their

knowledge in their daily work. One of the most

important developments concerning the nature of

tacit, collective knowledge in the contemporary

workplace has been the deployment of the concept

‘communities of practice (CoPs)’, by which we

mean groups of people with a common interest

where each member contributes knowledge about a

common domain (Wenger, 1998). CoPs is

necessarily bound to a technology, a set of

techniques or an organization, that is to a common

referent from which all members evaluate the

authority or skill and reputation of their peers and

the organization. A key factor for CoPs is provides

an environment of confidence where their members

can to share the information and best practices.

In order to provide to companies the conditions

to develop trustworthy knowledge management

systems we propose a multi-agent systems that

simulates the member’s behaviours of CoPs to detect

trustworthy knowledge sources. Thus in Section 2,

we explain why agents are a suitable technology

with which to manage knowledge. Then, in Section

3 we describe our proposal. After that, in Section 4

we illustrate how the multi-agent architecture has

been used to implement a prototype which detects

and suggests trustworthy knowledge sources for

members in CoPs. Finally, in Section 5 the

evaluation and future work are presented.

2 AGENTS IN KNOWLEDGE

MANAGEMENT

2.1 Why Intelligent Agents?

Due to the fundamentally social nature of knowledge

management applications different techniques have

been used to implement KMS. One of them, which

is proving to be quite useful is the agent paradigm

(van-Elst et al, 2003). Different definitions of

intelligent agents can be found in literature. For

instance, in (Mohammadian, 2004) agents are

defined as computer programs that assist users with

their tasks. One way of distinguishing agents from

other types of software applications and to

characterize them is to describe their main properties

(Wooldridge & Jennings, 1995):

 Autonomy: agents operate without the direct

intervention of humans or others, and have

some kind of control over their actions and

internal states.

 Social ability: agents interact with other agents

(and possibly humans) via some kind of agent

communication language.

 Reactivity: agents perceive their environment

and respond in a timely fashion.

 Pro-activeness: in the sense that the agents can

take the initiative and achieve their own goals.

In addition, intelligent agent’s specific

characteristics turn them into promising candidates

in providing a KMS solution (Mercer & Greenwood,

2001). Moreover, software agent technology can

monitor and coordinate events, meetings and

disseminate information (Balasubramanian et al,

2001), building and maintaining organizational

memories (Abecker et al, 2003). Another important

issue is that agents can learn from their own

experience. Most agents today employ some type of

artificial intelligence technique to assist the users

with their computer-related tasks, such as reading e-

mails, maintaining a calendar, and filtering

information. Agents can exhibit flexible behaviour,

providing knowledge both “reactively”, on user

request, and “pro-actively”, anticipating the user’s

knowledge needs. They can also serve as personal

assistants, maintaining the user’s profile and

preferences. The advantages that agent technology

has shown in the area of information management

have encouraged us to consider agents as a suitable

technique by which to develop an architecture with

the goal of helping to develop trustworthy KMS.

Therefore, we have chosen the agent paradigm

because it constitutes a natural metaphor for systems

with purposeful interacting agents, and this

abstraction is close to the human way of thinking

about their own activities (Wooldridge & Ciancarini,

2001). This foundation has led to an increasing

interest in social aspects such as motivation,

leadership, culture or trust (Fuentes et al, 2004). Our

research is related to this last concept of “trust” since

artificial agents can be made more robust, resilient

and effective by providing them with trust reasoning

capabilities.

2.2 Previous Work in the Field

This research can be compared with other proposals

that use agents and trust in knowledge exchange. For

instance, in (Abdul-Rahman & Hailes, 2000), the

authors propose a model that allows agents to decide

which agents’ opinions they trust more and propose

a protocol based on recommendations. This model is

based on a reputation or word-of-mouth mechanism.

The main problem with this approach is that every

ICSOFT 2007 - International Conference on Software and Data Technologies

220

agent must keep rather complex data structures that

represent a kind of global knowledge about the

whole network. In (Schulz et al, 2003), the authors

propose a framework for exchanging knowledge in a

mobile environment. They use delegate agents to be

spread out into the network of a mobile community

and use trust information to serve as the virtual

presence of a mobile user. Another interesting work

is (Wang & Vassileva, 2003) where the authors

describe a trust and reputation mechanism that

allows peers to discover partners who meet their

individual requirements through individual

experience and by sharing experiences with other

peers with similar preferences. This work is focused

on peer-to-peer environments.

Barber and Kim (2004) present a multi-agent

belief revision algorithm based on belief networks.

In their model the agent is able to evaluate incoming

information, to generate a consistent knowledge

base, and to avoid fraudulent information from

unreliable or deceptive information sources or

agents. This work has a similar goal to ours.

However, the means of attaining it are different. In

Barber and Kim’s case they define reputation as a

probability measure, since the information source is

assigned a reputation value of between 0 and 1.

Moreover, every time a source sends knowledge the

source should indicate the certainty factor that the

source has of that knowledge. In our case, the focus

is very different since it is the receiver who

evaluates the relevance of a piece of knowledge

rather than the provider as in Barber and Kim’s

proposal.

3 A THREE-LEVEL

MULTI-AGENT

ARCHITECTURE

Before defining our architecture it is necessary to

explain the conceptual model of an agent which, in

our case, is based on two related concepts: trust and

reputation. The former can be defined as confidence

in the ability and intention of an information source

to deliver correct information (Barber & Kim, 2004)

and the latter as the amount of trust an agent has in

an information source, created through interactions

with information sources. There are other definitions

of these concepts (Gambetta, 1988; Marsh, 1994).

However, we have presented the most appropriate

for our research since the level of confidence in a

source is, in our case, based upon previous

experience of this.

The reputation of an information source not only

serves as a means of belief revision in a situation of

uncertainty, but also serves as a social law that

obliges us to remain trustworthy to other people.

Therefore, people, in real life in general and in

companies in particular, prefer to exchange

knowledge with “trustworthy people” by which we

mean people they trust. People with a consistently

low reputation will eventually be isolated from the

community since others will rarely accept their

justifications or arguments and will limit their

interaction with them. It is for this reason that the

remainder of this paper deals mainly with reputation.

Figure 1: General architecture.

Taking these concepts into account we designed

a multi-agent architecture which is composed of

three levels (see Figure 1): reactive, deliberative and

social. The reactive and deliberative levels are

considered by other authors as typical levels that a

multi-agent system must have (Ushida et al, 1998).

The first level is frequently used in areas related to

robotics where agents react to changes in the

environment, without considering other processes

generated in the same environment. In addition, the

deliberative level uses a reasoning model in order to

decide what action to perform.

On the other hand, the last level (social) is not

frequently considered in an explicit way, despite the

fact that these systems (multi-agent systems) are

composed of several individuals, interactions

between them and plans constructed by them. The

social level is only considered in those systems that

try to simulate social behaviour. Since we wish to

emulate human feelings such as trust, reputation and

even intuition we have added a social level that

considers the social aspects of a community which

takes into account the opinions and behaviour of

each of the members of that community. Other

previous works have also added a social level. For

instance in (Imbert & de Antonio, 2005) the authors

try to emulate human emotions such as fear, thirst,

bravery, and also uses an architecture of three levels.

In the following paragraphs we will explain each

of these levels in detail.

Reactive level: This is the agent’s capacity to

perceive changes in its environment and to respond

to these changes at the precise moment at which they

AGENTS THAT HELP TO DETECT TRUSTWORTHY KNOWLEDGE SOURCES IN KNOWLEDGE

MANAGEMENT SYSTEMS

221

happen. It is in this level when an agent will execute

the request of another agent without any type of

reasoning. That is to say, the agent must act quickly

in the face of critical situations.

Deliberative level: The agent may also have a

behaviour which is oriented towards objectives, that

is, it takes the initiative in order to plan its

performance with the purpose of attaining its goals.

In this level the agent would use the information that

it receives from the environment, and from its

beliefs and intuitions, to decide which is the best

plan of action to follow in order to fulfill its

objectives.

Social level: This level is very important as our

agents are situated within communities and they

exchange information with other agents. Thanks to

this level they can cooperate with other agents. This

level represents the actual situation of the

community, and also considers the goals and

interests of each community member in order to

solve conflicts and problems which may arise

between them. In addition, this level provides the

support necessary to measure and stimulate the level

of participation of the members of the community.

Two further important components of our

architecture are the Interpreter and the Planner (see

Figure 2). The former is used to perceive the

changes that take place in the environment. The

planner indicates how the actions should be

executed.

In the following subsections we will describe

each of the levels of which our architecture is

composed in more detail.

3.1 Reactive Architecture

This architecture was designed to the reactive level

of the agent. The architecture must respond at the

precise moment in which an event has been

perceived. For instance when an agent is consulted

about its position within the organization. This

architecture is formed of the following modules:

Figure 2: Reactive architecture.

Agent’s internal model: As a software agent

represents a person in a community this model stores

the user’s features. Therefore, this module stores the

following parts:

- The interests. This part is included in the

internal model in order to make the

process of distributing knowledge as fast

as possible. That is, the agents are able to

search for knowledge automatically,

checking whether there is stored

knowledge which matches with its own

interests. This behaviour fosters

knowledge sharing and reduces the

amount of work employees have to do

because they receive knowledge without

making searches.

- Expertise. This term can be briefly defined

as the skill or knowledge of a person who

knows a great deal about a specific thing.

Since we are emulating communities of

practice it is important to know the degree

of expertise that each member of the

community has in order to decide how

trustworthy a piece of knowledge is, since

people often trust in experts more than in

novice employees.

- Position. Employees often consider

information that comes from a boss as

being more reliable than that which comes

from another employee in the same (or a

lower) position as him/her (Wasserman &

Glaskiewics, 1994). Such different

positions inevitably influence the way in

which knowledge is acquired, diffused and

eventually transformed in the local area.

Because of this these factor will be

calculated in our research by taking into

account a weight that can strengthen this

factor to a greater or to a lesser degree.

- Profile. This part is included in the internal

model to describe the profile of the person

on whose behalf the agent is acting.

Therefore, a person’s preferences are

stored here.

Behaviour generator: This component is

necessary for the development of this architecture

since it has to select the agent’s behaviour. This

behaviour is defined on the basis of the agent’s

beliefs. Moreover, this component finds an

immediate response to the perceptions received of

the environment.

History: This component stores the interactions

of the agents with the environment.

Belief generation: This component is one of the

most important in the cognitive model because it is

in charge of creating and storing the agent’s

knowledge. Moreover, it defines the agent’s beliefs.

ICSOFT 2007 - International Conference on Software and Data Technologies

222

Beliefs: The beliefs module is composed of three

kinds of beliefs: inherited beliefs, lessons learned

and interactions. Inherited beliefs are the

organization’s beliefs that the agent receives. For

instance: an organizational diagram of the enterprise,

the philosophy of the company or community.

Lessons learned are the lessons that the agent

obtains while it interacts with the environment. The

information about interactions can be used to

establish parameters in order to know which the

agent can trust (agents or knowledge sources). This

module is based on the interests and goals of the

agent, because each time a goal is realised, the

lessons and experiences generated to attain this goal

are introduced in the agent’s beliefs as lessons

learned.

3.2 Deliberative Architecture

This architecture was designed to the deliberative

level of the agent (see Figure 3).

Figure 3: Deliberative architecture

Its components are:

Agent’s internal model: this module is the same

as that which is described in the reactive

architecture. It is composed of the interests, profile,

position and expertise of the agent.

Plans processor: This module is the most

important of this architecture as it is in charge of

evaluating the beliefs and goals to determine which

plans have to be included in the Planner to be

executed.

Belief generator: This component, as in the

previous architecture, is in charge of creating,

storing and retaining the agent’s knowledge. In

addition, it is also in charge of establishing the

agent’s beliefs. The belief creation process is a

continuous process that is initiated at the moment at

which the agent is created and which continues

during its entire effective life.

Intuitions: Intuitions are beliefs that have not

been verified but which it thinks may be true.

According to (Mui et al, 2002) intuition has not yet

been modelled by agent systems. In this work we

have tried to adapt this concept because we consider

that in real communities people are influenced by

their intuitions when they have to make a decision or

believe in something. This concept is emulated by

comparing the agents’ profiles to obtain an initial

value of intuition that can be used to form a belief

about an agent.

History: This component stores the interactions

of the agents with the environment.

3.3 Social Architecture

This architecture (see Figure 4) is quite similar to the

deliberative architecture. The main differences are

the social model and social behaviour processor,

which are explained in the following paragraphs.

Figure 4: Social architecture.

Social model: This module represents the actual

state of the community, the community’s interests

and the members’ identifiers.

Social behaviour processor: This component

processes the beliefs of the community’s members.

To do this, this module needs to manage the goals,

intuitions and beliefs of the community in order to

make a decision.

The social focus that this architecture provides

permits us to give the agents the social behaviour

necessary to emulate the work relationships in an

organization. In addition, this layer permits the

decentralization of decision making, that is, it

provides methods by which to process or make

decisions based on the opinions of the members of a

community.

4 IMPLEMENTATION OF THE

ARCHITECTURE

To evaluate the feasibility of the implementation of

the architecture, we have developed a prototype into

which people can introduce documents and where

these documents can also be consulted by other

people. The goal of this prototype is to allow

software agents to help employees to discover the

information that may be useful to them thus

decreasing the overload of information that

employees often have and strengthening the use of

AGENTS THAT HELP TO DETECT TRUSTWORTHY KNOWLEDGE SOURCES IN KNOWLEDGE

MANAGEMENT SYSTEMS

223

knowledge bases in enterprises. In addition, we try

to avoid the situation of employees storing valueless

information in the knowledge base.

A feature of this system is that when a person

searches for knowledge in a community, and after

having used the knowledge obtained, that person

then has to evaluate the knowledge in order to

indicate whether:

 The knowledge was useful.

 How it was related to the topic of the search (for

instance a lot, not too much, not at all).

Figure 5: Agent’s distribution.

One type of agent in our prototype (see Figure 5)

is the User Agent which is in charge of representing

each person that may consult or introduce

knowledge in a knowledge base. The User Agent can

assume three types of behavior or roles similar to the

tasks that a person may carry out in a knowledge

base. Therefore, the User Agent plays one role or

another depending upon whether the person that it

represents carries out one of the following actions:

 The person contributes new knowledge to the

communities in which s/he is registered. In this

case the User Agent plays the role of Provider.

 The person uses knowledge previously stored in

the community. Then, the User Agent will be

considered as a Consumer.

 The person helps other users to achieve their

goals, for instance by giving an evaluation of

certain knowledge. In this case the role is that of

a Partner. So, Figure 5 shows that in

Community 1 there are two User Agents

playing the role of Partner (Pa), one User Agent

playing the role of Consumer (Co) and another

being a Provider (Pr).

The second type of agent within a community is

called the Manager Agent (represented in black in

Figure 5) which is in charge of managing and

controlling its community. In order to approach this

type of agent the following tasks are carried out:

 Registering an agent in its community. It thus

controls how many agents there are and how

long the stay of each agent in that community is.

 Registering the frequency of contribution of

each agent. This value is updated every time an

agent makes a contribution to the community.

 Registering the number of times that an agent

gives feedback about other agents’ knowledge.

For instance, when an agent “A” uses

information from another agent “B”, the agent

A should evaluate this information. Monitoring

how often an agent gives feedback about other

agents’ information helps to detect whether

agents contribute to the creation of knowledge

flows in the community since it is as important

that an agent contributes new information as it

is that another agent contributes by evaluating

the relevance or importance of this information.

 Registering the interactions between agents.

Every time an agent evaluates the contributions

of another agent the Manager agent will register

this interaction. But this interaction is only in

one direction, which means, if the agent A

consults information from agent B and evaluates

it, the Manager records that A knows B but that

does not means that B knows A because B does

not obtain any information about A.

Moreover, when a user wants to join to a

community in which no member knows anything

about him/her, the reputation value assigned to the

user in the new community is calculated on the basis

of the reputation assigned from others communities

where the user is or was a member. For instance, an

User Agent called j, will ask each community

manager where he/she was previously a member to

consult each agent which knows him/her with the

goal of calculating the average value of his/her

reputation (R

). This is calculated as:

(1)

where n is the number agents who know j and R

is the value of reputation of j in the eyes of s. In the

case of being known in several communities the

average of the values R

will be calculated. Then,

the User Agent j presents this reputation value

(similar to when a person presents his/her

curriculum vitae when s/he wishes to join a

company) to the Manager Agent of the community

to which it is “applying”. This reputation value

permits to assign a reputation value taking into

account the previous experiences and relations with

others agents, generating a flow and exchange of

information between the agents. This mechanism is

similar to the “word-of-mouth” propagation of

information for a human (Abdul-Rahman & Hailes,

2000).

ICSOFT 2007 - International Conference on Software and Data Technologies

224

In addition, R

value is computed as follows:

(2)

where E

is the value of expertise which is

calculated according to the degree of experience that

the person upon whose behalf the agent acts has in a

domain.

is the value assigned to a person’s position.

This position is defined in the agent’s internal model

of the reactive architecture described in Section 3.1.

is the value assigned to intuition which is

calculated by comparing each user’s profile.

Intuition is an important component both in the

deliberative and in the social architecture because it

helps agents to create their beliefs and behavior

according to their own features.

In addition, previous experience should also be

calculated. We suppose that when an agent A

consults information from another agent B, the agent

A should evaluate how useful this information was.

This value is called QC

(Quality of j’s

Contribution). To attain the average value of an

agent’s contribution, we calculate the sum of all the

values assigned to these contributions and we divide

it between their total. In the expression n represents

the total number of evaluated contributions.

Finally, w

, w

and w

are weights with which the

Reputation value can be adjusted to the needs of the

organizations or communities. These weights

represent different values depending on the category

of each employee. For instance, if an enterprise

considers that all its employees have the same

category, then w

=0. The same could occur when the

organization does not take its employee’s intuitions

or expertise into account.

In this way, an agent can obtain a value related to

the reputation of another agent and decide to what

degree it is going to consider the importance of the

information obtained from this agent. The formulas

(1) and (2) are processed in the social and

deliberative architecture respectively.

5 EVALUATION AND FUTURE

WORK

Once the prototype has finished we will evaluate it.

To do this, different approaches can be followed,

from a multi-agent point of view or from a social

one. First of all we have focused on the former and

we are testing the most suitable number of agents

advisable for a community. Therefore, several

simulations have been performed. As result of them

we found that:

- The maximum number of agents supported by

the Community Manager Agent when it

receives User Agents’ evaluations is

approximately 800. When we tried to work

with 1000 agents for instance, the messages

were not managed conveniently. However, we

could see that the Manager Agent could

support a high number of petitions, at least,

using simpler behavior.

- On the other hand, if we have around 10 User

Agents launched, they need about 20 o more

interactions to know all agents of the

community. If a User Agent has between 10

and 20 interactions with other members it is

likely that it interacts with 90% of members of

its community, which means that the agent is

going to know almost all the members of the

community. Therefore, after several trials we

detected that the most suitable number of

agents for one community was around 10

agents and they needed a average of 20

interactions to know (to have a contact with)

all the members of the community, which is

quite convenient in order to obtain its own

value of reputation about other agent.

All these results are being used to detect whether

the exchange of messages between the agents is

suitable, and to see if the information that we

propose to be taken into account to obtain a

trustworthy value of the reputation of each agent is

enough, or if more parameters should be considered.

Once this validation is finished we need to carry out

further research to answer one important question,

which is how the usage of this prototype affects the

performance of a community. This is the social

approach that we mentioned at the beginning of this

section. As claimed in (Geib et al, 2004) to measure

the performance of communities is a challenge since

communities only have an indirect impact on

business results. In order to do this we are going to

take some ideas of the performance measurement

framework for communities propose by

(McDermott, 2002) where the performance of

communities is measured in terms of output and

values such as: personal knowledge, strength of

relationships (this could be one of the most

important values for our research) and access to

information. This research will be critical to find

how our proposal affects communities of practice.

ACKNOWLEDGEMENTS

This work is partially supported by the ENIGMAS

(PIB-05-058), and MECENAS (PBI06-0024)

AGENTS THAT HELP TO DETECT TRUSTWORTHY KNOWLEDGE SOURCES IN KNOWLEDGE

MANAGEMENT SYSTEMS

225

project. It is also supported by the ESFINGE project

(TIN2006-15175-C05-05) Ministerio de Educación

y Ciencia (Dirección General de Investigación)/

Fondos Europeos de Desarrollo Regional (FEDER)

in Spain

and CONACYT (México) under grant of

the scholarship 206147 provided for the first author.

REFERENCES

Abdul-Rahman, A., Hailes, S., (2000),Supporting Trust in

Virtual Communities. 33rd Hawaii International

Conference on Systems Sciences (HICSS'00).

Abecker, A., Bernardi, A., van-Elst, L., (2003), Agent

Technology for Distributed Organizational Memories:

The Frodo Project. In Proceedings of 5th

International Conference on Enterprise Information

Systems. Angers, France.

Alavi, M., Leidner, D. E., (2001), Knowledge

Management and Knowledge Management Systems:

Conceptual Foundations and Research Issues. MIS

Quarterly, Vol. 25, No. 1, pp. 107-136.

Balasubramanian, S., Brennan, R., Norrie, D., (2001), An

Architecture for Metamorphic Control of Holonic

Manufacturing Systems. Computers in Industry, Vol.

46(1), pp. 13-31.

Barber, K., Kim, J., (2004), Belief Revision Process Based

on Trust: Simulation Experiments. In 4th Workshop on

Deception, Fraud and Trust in Agent Societies,

Montreal Canada.

Fuentes, R., Gómez-Sanz, J., Pavón, J., (2004), A Social

Framework for Multi-agent Systems Validation and

Verification, in Wang, S. et al (Eds.) ER Workshop

2004, Springer Verlag, LNCS 3289, pp. 458-469.

Gambetta, D., (1988), Can We Trust Trust? In D.

Gambetta, editor, Trust: Making and Breaking

Cooperative Relations, pp. 213-237.

Geib, M., Braun, C., Kolbe, L., Brenner, W., (2004),

Measuring the Utilization of Collaboration

Technology for Knowledge Development and

Exchange in Virtual Communities. Proceedings of the

37th Annual Hawaii International Conference on

System Sciences (HICSS'04). IEEE Computer Society

Press, Vol. 1.

Huysman, M., Wit, D., (2000), Knowledge Sharing in

Practice. Kluwer Academic Publishers, Dordrecht.

Imbert, R., and de Antonio, A.,(2005), When emotion does

not mean loss of control. Lecture Notes in Computer

Science, T. Panayiotopoulos, J. Gratch, R. Aylett, D.

Ballin, P. Olivier, and T. Rist, Eds. Springer-Verlag,

London., pp. 152-165.

Desouza, K, Awazu, Y., Baloh, P., (2006), Managing

Knowledge in Global Software Development Efforts:

Issues and Practices. IEEE Software, Vol. 23, No. 5,

pp. 30-37.

Lawton, G., (2001), Knowledge Management: Ready for

Prime Time? Computer, Vol. 34, No. 2, pp. 12-14.

Marsh, S., (1994), Formalising Trust as a Computational

Concept, PhD Thesis, University of Stirling.

McDermott, R., (2002), Measuring the Impact of

Communities. Knowledge Management, 2002. Vol. 5,

No. 2, pp. 26-29.

Mercer, S., Greenwood, S., (2001), A Multi-Agent

Architecture for Knowledge Sharing. In Proceedings

of the Sixteenth European Meeting on Cybernetics and

Systems Research.

Mohammadian, M., (2004), Computational Intelligence

Techniques Driven Intelligent Agents for Web Data

Mining and Information Retrieval, in Intelligent

Agents for Data Mining and Information Retrieval.

IDEA Group Publishing. pp. 15-29.

Mui, L., Halberstadt, A., Mohtashemi, M., (2002), Notions

of Reputation in Multi-Agents Systems: A Review,

International Conference on Autonomous Agents and

Multi-Agents Systems (AAMAS'02), pp. 280-287.

Nebus, J., (2001), Framing the Knowledge Search

Problem: Whom Do We Contact, and Why Do We

Contact Them? Academy of Management Best Papers

Proceedings, pp. 1-7.

Schulz, S., Herrmann, K., Kalcklosch, R., Schowotzer, T.,

2003, Trust-Based Agent-Mediated Knowledge

Exchange for Ubiquitous Peer Networks. AMKM,

LNAI 2926, p. 89-106.

Ushida, H., Hirayama, Y., Nakajima, H., (1998) Emotion

Model for Life like Agent and its Evaluation. In

Proceedings of the Fifteenth National Conference on

Artificial Intelligence and Tenth Innovative

Applications of Artificial Intelligence Conference

(AAAI'98 / IAAI'98). Madison, Wisconsin, USA.

van-Elst, L., Dignum, V., Abecker, A., (2003) Agent-

Mediated Knowledge Management. in International

Simposium AMKM'03, Stanford, CA, USA: Springer.

Wang, Y., Vassileva, J., (2003), Trust and Reputation

Model in Peer-to-Peer Networks. Proceedings of IEEE

Conference on P2P Computing.

Wasserman, S. and Glaskiewics, J., (1994), Advances in

Social Networks Analysis. Sage Publications.

Wenger, E., (1998), Communities of Practice: Learning

Meaning, and Identity, Cambridge U.K.: Cambridge

University Press.

Wooldridge, M. and Ciancarini, P., (2001), Agent-

Oriented Software Engineering: The State of the Art.

In Wooldridge M., Ciancarini, P. (Eds.), Agent

Oriented Software Engineering. Springer Verlag,

LNAI 1975.

Wooldridge, M., Jennings, N.R., (1995), Intelligent

Agents: Theory and Practice. Knowledge Engineering

Review, Vol. 10, No. 2, pp. 115-152.

Zand, D., (1997), The Leadership Triad: Knowledge,

Trust, and Power. Oxford University Press.

ICSOFT 2007 - International Conference on Software and Data Technologies

226