MULTI-AGENT SYSTEMS IN INTELLIGENT PERVASIVE

SPACES

Joaquim Filipe

Escola Superior de Tecnologia de Setúbal, Instituto Politécnico de Setúbal

Rua Vale de Chaves, Estefanilha, 2910-761 Setúbal, Portugal

Keywords: Multi-agent Systems, Theoretical Agent Models, Organisational Semiotics, Social Contexts.

Abstract: This paper describes an agent model based on social psychology and also on the concept of organisational

semiotics information fields to provide a conceptual infrastructure for designing multi-agent systems in

intelligent pervasive spaces. Since ‘information’ is an ill-defined word we prefer to adopt the semiotics

framework, which uses the ‘sign’ as the elementary concept. Information as a composition of signs is then

analyzed at different levels, including syntax, semantics, pragmatics and the social level. Based on different

properties of signs, found at different semiotic levels, we adopt the EDA agent model (an acronym for its

three component modules: Epistemic-Deontic-Axiological). Intelligent pervasive spaces are ICT-enhanced

physical and social spaces, differing from traditional pervasive computing on the focus, which in pervasive

spaces is essentially social instead of technological (Liu et al., 2010). Agents are often described in terms of

their internal structure, emphasizing their autonomy even in social settings involving communication and

coordination. In this paper we suggest that agents can be seen both as individual and social entities,

simultaneously. The norm-based multi-agent social architecture defined in this paper is flexible enough to

accommodate changes in social structure, including changes in role specification, and representation of

inter-subjective social objects.

1 INTRODUCTION

Social groups can be seen as multi-agent systems,

possibly including both human and artificial agents.

If there is a strong social cohesiveness, then we may

be in the presence of organisations, which can be

modeled as multilayered Information Systems (IS),

composed of an informal subsystem, a formal

subsystem and a technical system as shown in figure

1. this structure is typical of the organisational

semiotics perspective on information systems.

Organisational Semiotics is a particular branch of

Semiotics, the formal doctrine of signs (Peirce,

1931-1935), concerned with analyzing

and modeling organisations as information systems.

Core concepts such as information and

communication are very complex and ill-defined

concepts, which should be analysed in terms of more

elementary notions such as semiotic ‘signs’.

Business processes would then be seen as processes

involving the creation, exchange and use of signs.

Since organisational activity is an information

process based on the notion of responsible co-

operative agents, we propose a model that

accommodates both the social dimension in

organisational agents behavior and the relative

autonomy that individual agents exhibit in real

organisations. The proposed model is an intentional

model, based on three main components, each of

them trying to capture particular relevant agent

attitudes. Agents are seen as intelligent units of a

larger distributed system, in the sense that each unit

has an autonomous capacity to infer and act, based

on a knowledge-based infrastructure (Filipe, 2002).

In our research work intelligent agents are placed

in social information fields, or spaces, where they

interact with other artificial agents or humans, on

behalf of human users or human organisations, who

must ultimately take responsible for the behavior of

each artificial agent.

Organisational Semiotics, however, is not

sufficiently developed to provide an analytical

Filipe J.

MULTI-AGENT SYSTEMS IN INTELLIGENT PERVASIVE SPACES.

DOI: 10.5220/0003258800090016

In Proceedings of the Twelfth International Conference on Informatics and Semiotics in Organisations (ICISO 2010), page

ISBN: 978-989-8425-26-3

Figure 1: Three main layers of the real information system (Stamper 1996).

model for designing each agent in the organisational

social system. This paper extends the work that has

been done in semantic analysis, providing a way to

clearly define the specification of individual agents

at a pragmatic level, keeping a social and normative

perspective.

2 ORGANISATIONAL

SEMIOTICS

In this paper we approach the problem of

constructing multi-agent systems in pervasive spaces

using the Organisational Semiotics stance (Stamper,

1973; Liu, 2000), to provide adequate system

requirements and a solid conceptual basis.

Semiotics, traditionally divided into three areas –

syntax, semantics and pragmatics – has been

extended by Stamper in order to incorporate three

other levels, including a social world level. A

detailed and formal account of these levels may be

found in (Stamper, 1996).

This approach is different from mainstream

computer science because instead of adopting an

objectivist stance – where it is assumed the existence

of a single observable reality, external to the agent,

which some modeling methods try to capture with

the help of some software engineering approach – it

adopts a social subjectivist stance. This means that

for all practical purposes nothing exists without a

perceiving agent nor without an agent engaging in

action (Stamper, 2000). Invariant behaviors

available to an agent are called affordances. This

philosophical stance ties every item of knowledge to

an agent who is, in a sense, responsible for it.

The recent paradigm shift from centralized data

processing architectures to heterogeneous distributed

computing architectures, emerging especially since

the 1990’s, placed social concerns in the agenda of

much research activity in Computing, particularly in

the Distributed Artificial Intelligence field (DAI). In

DAI, organisations are modeled as multi-agent

systems composed by autonomous agents acting in

order to achieve social goals, in a cooperative

manner (Wooldridge and Jennings, 1995; Singh,

1996; Filipe, 2000). Social goals can be seen as

norms therefore we hypothesize that the

conceptualization and development of these

intelligent pervasive systems require normative

models.

3 THE EDA MODEL

Social psychology provides a well-known

classification of norms, partitioning them into

perceptual, evaluative, cognitive and behavioral

norms. These four types of norms are associated

with four distinct attitudes, respectively (Stamper et

al., 2000):

• Ontological – to acknowledge the existence

of something (related to perception);

• Axiological – to be disposed in favor or

against something in value terms;

• Epistemic – to adopt a degree of belief or

disbelief;

• Deontic – to be disposed to act in some way.

Our agent model is based on these attitudes and the

associated norms, which we characterize in more

detail below:

• Perceptual norms, guided by evaluative

norms, determine what signs the agent

chooses to perceive. Then, when a sign is

perceived, a pragmatic function will update

the agent EDA model components

accordingly.

• Cognitive norms define entity structures,

semantic values and cause-effect

relationships, including both beliefs about the

INFORMAL IS: a sub-culture where meanings are established, intentions are understood,

beliefs are formed and commitments with responsibilities are made, altered and discharged

FORMAL IS: bureaucracy where form and rule replace meaning and intention

TECHNICAL IS: Mechanisms to automate part of the formal

system

ICISO 2010 - International Conference on Informatics and Semiotics in Organisations

present state and expectations for the future.

Conditional beliefs are typically represented

by rules, which being normative allow for the

existence of exceptions.

• Behavioral norms define what an agent is

expected to do. These norms prescribe ideal

behaviors as abstract plans to bring about

ideal states of affairs, thus determining what

an agent ought to do. Deontic logic is a

modal logic that studies the formal properties

of normative behaviors and states.

• Evaluative norms are required for an agent to

choose its actions based on both epistemic

and deontic attitudes. If we consider a

rational agent, then the choice should be such

that the agent will maximize some utility

function, implicitly defined as the integral of

the agent’s axiological attitudes.

Figure 2: The EDA agent model.

Using this taxonomy of norms, and based on the

assumption that an organisational agent behavior is

determined by the evaluation of deontic norms given

the agent epistemic state

, we propose an intentional

agent model, which is decomposed into three

components: the epistemic, the deontic and the

axiological (Figure 2).

Together, these components incorporate all the

agent informational contents, where it is shown that

information is a complex concept, and requires

different viewpoints to be completely analyzed. The

description and detailed analysis of each of the

aforementioned components is provided in (Filipe

and Liu, 2000).

4 RELATED WORK

Although inspired mainly in the semiotics stance,

and the norms-attitudes relationships at different

psycho-sociological levels, related to organisational

modeling, the EDA agent model is related to several

von Wright (1968) suggests that the study of deontic

concepts and the study of the notions of agency and

activity are intertwined.

other models previously proposed, mainly in the

DAI literature.

One of these is the BDI model (Belief, Desire,

Intention) proposed by Rao and Georgeff (1991).

This model is based on a theory of intentions,

developed by Bratman (1987). The BDI perspective

is more concerned with capturing the properties of

human intentions, and their functions in human

reasoning and decision making, whereas the EDA

model is a norm-based representation of beliefs,

goals and values, based on a semiotics view of

information and oriented towards understanding and

modeling social cooperation. BDI agents can easily

abstract from any social environment because they

are not specifically made for multi-agent systems

modeling.

Singh (1996) also provides a social perspective to

multi-agent systems. He adopts a notion of

commitment that bears some similarity with our

goals, in the sense that it relates a proposition to

several agents, defining the concept of ‘sphere of

commitment’.

Jennings (1994) proposes a social coordination

mechanism based on commitments and conventions,

supported by the notions of joint beliefs and joint

intentions.

Yu and Mylopoulos (1997) also recognized the

importance of explicitly representing and dealing

with goals, in terms of means-ends reasoning, and

they have proposed the i* modeling framework, in

which organisations and business process models are

based on dependency relationships among agents.

5 INTENTIONS AND SOCIAL

NORMS IN THE PERVASIVE

SPACES

Based on this agent model, we can create social

structures composed of many interacting agents. The

multi-agent system metaphor that we have adopted

for modelling organisations implies that

organisations are seen as goal-governed collective

agents, which are composed of individual agents.

This perspective comes in line with the principles of

normative agents proposed in (Castelfranchi, 1993).

The social normative structure is essentially

defined by agent roles and relationships. Roles can

then be instantiated by one or more agents.

Conceptually, a role is a set of Services and Policies,

and a Policy is a set of Obligations and

Authorizations. At the implementation level, agents

are represented by objects and services are defined

MULTI-AGENT SYSTEMS IN INTELLIGENT PERVASIVE SPACES

by the object interface. Policies are sets of rules

related to one or more EDA components, each of

which includes at least one knowledge base (KB).

When an agent is selected to perform a role, each of

its EDA components downloads the adequate KB

from an organisational role server.

Obligations are represented as particular goals

whereas authorizations are represented using the

same syntax as goals but in a pattern format, and are

Figure 3: Social and Individual goals parallelism in the

EDA model.

interpreted as potential action enabling/blocking

devices.

Figure 3 describes the parallelism between mental

and social constructs that lead to setting a goal in the

agent's agenda, and which justifies the adoption of

an obligation. Here, p represents a proposition

(world state).

()

represents p as one of agent

’s beliefs. ()Op

represents the obligation that

must see to it that p is true for

. ( )Op

represents the interest that

has on seeing to it that

p is true for itself – a kind of self-imposed

obligation. In this diagram

(,)

∈Ε

means,

intuitively, that proposition p is one of the goals on

’s agenda.

Interest is one of the key notions that are

represented in the EDA model, based on the

combination of the deontic operator ‘ought-to-be’

(von Wright, 1951) and the agentive ‘see-to-it-that’

stit operator (Belnap, 1991).

Interests and Desires are manifestations of

Individual Goals. The differences between them are

the following:

• Interests are individual goals of which the agent

is not necessarily aware, typically at a high

abstraction level, which would contribute to

improve its overall utility. Interests may be

originated externally, by other agents’

suggestions, or internally, by inference:

deductively (means-end analysis), inductively or

abductively. One of the most difficult tasks for

an agent is to become aware of its interest areas

because there are too many potentially

advantageous world states, making the full utility

evaluation of each potential interest impossible,

given the limited reasoning capacity of any

agent.

• Desires are interests that the agent is aware of.

However, they may not be achievable and may

even conflict with other agent goals; the logical

translation indicated in the figure,

() ( ())Op BOp

αα

ααα

∧ , means that desires are

goals that agent

ought to pursue for itself and

that it is aware of. However, the agent has not yet

decided to commit to it, in a global perspective,

i.e. considering all other possibilities. In other

words, desires become intentions only if they are

part of the preferred extension of the normative

agent EDA model (Filipe, 2000).

It is important to point out the strong connection

between these deontic concepts and the axiologic

component. All notions indicated in the figure

should be interpreted from the agent perspective, i.e.

values assigned to interests are determined by the

agent. Eventually, external agents may consider

some goal (interest) as having a positive value for

the agent and yet the agent himself may decide

otherwise. That is why interests are considered here

to be the set of all goals to which the agent would

assign a positive utility, but which it may not be

aware of. In that case the responsibility for the

interest remains on the external agent.

Not all interests become desires but all desires are

agent interests. This may seem contradictory with a

situation commonly seen in human societies of

agents acting in others’ best interests, sometimes

even against their desires: that’s what parents do for

their children. However, this does not mean that the

agent desires are not seen as positive by the agent; it

only shows that the agent may have a deficient

axiologic system (by its information field standards)

and in that case the social group may give other

Individual

Goal

Social

Obligation

Interest



()Op

Duty



()Op

Origin:

-Other agents

-Inference

Origin:

-Social roles

-Inference

Desire



() ())(Op BOp

αα

∧

Demand



() ())(Op BOp

ββ

αα

∧

Intention



(,)

∈Ε

Agenda

Achievement

External Internal(inEDA)

Awareness

ICISO 2010 - International Conference on Informatics and Semiotics in Organisations

agents the right to override that agent. In the case of

artificial agents such a discrepancy would typically

cause the agent to be banned from the information

field (no access to social resources) and eventually

repaired or discontinued by human supervisors, due

to social pressure (e.g. software viruses).

In parallel with Interests and Desires, there are

also social driving forces converging to influence

individual achievement goals, but through a different

path, based on the general notion of social

obligation. Social obligations are the goals that the

social group where the agent is situated require the

agent to attain. These can also have different

flavours in parallel to what we have described for

individual goals.

• Duties are social goals that are attached to the

particular roles that the agent is assigned to,

whether the agent is aware that they exist or not.

The statement

()Op

means that agent

ought to do p on behalf of another agent

Agent

may be another individual agent or a

collective agent, such as the society to which

belongs. Besides the obligations that are

explicitly indicated in social roles, there are

additional implicit obligations. These are inferred

from conditional social norms and typically

depend on circumstances. Additionally, all

specific commitments that the agent may agree

to enter also become duties; however, in this

case, the agent is necessarily aware of them.

• Demands are duties that the agent is aware of

This notion is formalised by the following

logical statement:

() ( ())Op BOp

ββ

ααα

∧

. Social

demands motivate the agent to act but they may

not be achievable and may even conflict with

other agent duties; being autonomous, the agent

may also decide that, according to circumstances,

it is better not to fulfill a social demand and

rather accept the corresponding sanction.

Demands become intentions only if they are part

of the preferred extension of the normative agent

EDA model – see (Filipe, 2000 section 5.7) for

details.

• Intentions: Whatever their origin (individual or

social) intentions constitute a non-conflicting set

of goals that are believed to offer the highest

possible value for the concerned agent.

Intentions are designated by some authors

According to the Concise Oxford Dictionary, demand is

“an insistent and peremptory request, made as of right”.

We believe this is the English word with the closest

semantics to what we need.

(Singh, 1990) as psychological commitments (to

act). However, intentions may eventually

(despite the agent sincerity) not actually be

placed in the agenda, for several reasons:

o They may be too abstract to become directly

executed, thus requiring further means-end

analysis and planning.

o They may need to wait for their appropriate time

of execution.

o They may be overridden by higher priority

intentions.

o Required resources may not be ready.

6 INFORMATION FIELDS FOR

INTER-SUBJECTIVE

REPRESENTATION OF SOCIAL

OBJECTS

Following Habermas (1984) we postulate the

existence of a shared ontology or inter-subjective

reality that defines the social context (information

field) where agents are situated (Filipe, 2003). This

kind of social shared knowledge is not reducible to

individual mental objects (Conte and Castelfranchi,

1995). For example, in the case of a commitment

violation, sanction enforcement is explicitly or

tacitly supported by the social group to which the

agents belong, otherwise the stronger agent would

have no reason to accept the sanction. This

demonstrates the inadequacy of the reductionist

view.

Once again, we look at human organisational

models for designing multi-agent systems; for

example, contracts in human societies are often

written and publicly registered in order to ensure the

existence of socially accepted, and trusted, witnesses

that would enable the control of possible violations

at a social level. Non-registered contracts and

commitments are often dealt with at a bilateral level

only and each concerned agent has its internal

contract copy. This observation suggests two

representational models:

• A distributed model: Every agent keeps track of

social objects in which that agent is involved

and may also be a witness of some social

objects involving other agents.

• A centralised model: There is an Information

Field Server (IFS) that has a social objects

database, including shared beliefs, norms, agent

roles, social commitments, and institutions.

The distributed model is more robust to failure,

given the implicit redundancy. For example, a

MULTI-AGENT SYSTEMS IN INTELLIGENT PERVASIVE SPACES

Figure 4: Social objects representation at inter-subjective level and their usage.

contract where a number of parties are involved is

kept in all concerned agents’ knowledge bases,

therefore if an agent collapses the others can still

provide copies of the contract. It is also more

efficient assuming that all agents are honest and

sincere; for example, commitment creation and

termination involved in business transactions would

not need to be officially recorded – a simple

representation of a social commitment at the

concerned agents EDA model would suffice.

However, since these assumptions are often

unrealistic, the distributed model cannot completely

replace the role of certified agents, trusted by society

to keep record of shared beliefs and social

commitments. We assume here that these social

notions are part of the ontology that is shared by all

members of an information field; that’s why we call

these trusted repositories of the shared ontology

“Information Field Servers”. These servers have the

following characteristics:

• Different information fields must have different

IFS because the shared ontology may differ

among specific information fields.

• Each information field may have several non-

redundant IFS, each representing a small part of

the shared ontology.

• The robustness problems of IFS are minimized

by reliable backup (redundant) agents.

Considering the empirical semiotics level,

communication bandwidth is another relevant factor

to consider: if all social objects were placed in

central IFS agents these might become system

bottlenecks.

A conceptual problem that exists but is not in the

scope of this paper is related to the representation of

social objects resulting from the interaction of agents

belonging to different information fields. Possible

solutions range from the unification of the different

conceptual frameworks to the creation of new

information fields where the ontology is constructed

from a continuous meaning negotiation process via

the interaction of the concerned agents.

In figure 4 the architecture of the inter-subjective

level is depicted with respect to the localisation of

social objects in addition to an example showing

how social objects are used at the subjective level.

Commitments are first class objects, which can be

represented either in the agents’ EDA models (which

we designate as the agents’ space) or in the IFS’

EDA model (which we designate as the Information

Field Server’s space). In the example above, agents

A1 and A2 have only an internal representation (in

each EDA model) of a shared commitment C1,

whereas Agents A2 and A3 do not have an internal

representation of commitment C2 because this

commitment is represented in IFS1. All agents A2

and A3 need is a reference (i.e. a pointer) to that

shared commitment, although for implementation

reasons related to communication bandwidth and

efficiency copies may be kept internally.

7 CONCLUSIONS AND FUTURE

WORK

The EDA model described here is based on the

organisational semiotics stance, where normative

knowledge and norm-based coordination is

emphasized. The main model components

(Epistemic, Deontic and Axiological) reflect a social

Inter‐subjectiveLevelObjectsLocalisation



Regularagents’space   InformationFieldServers’space

      

IFS



SubjectiveLevel

CommitmentC1 CommitmentC2 

AgentA1 AgentA2 AgentA3

ICISO 2010 - International Conference on Informatics and Semiotics in Organisations

psychology classification of norms, therefore

provide a principled norm-based structure for the

agent internal architecture that is also oriented

toward a norm-based social interaction in

organisations.

The EDA architecture integrates a number of

ideas gathered from the DAI field and from deontic

logic. Some of the most important ones were

described in the previous section. We recognize the

need for a semantics to underpin the proposed model

but, at the present, we have focused mainly on

conceptual issues.

Particularly important for social modeling is the

notion of ‘commitment’. Although we didn’t

formally define our notion of commitment, we do

see commitments in terms of goals, emerging as a

pragmatic result of social interaction. We believe

that multi-agent commitments can be modeled as

related sets of deontic-action statements, distributed

across the intervening agents, based on the notion of

unified goals as proposed in the deontic component

of our model.

An axiological component seems to be a

necessary part of any intelligent agent, both to

establish preferred sets of agent beliefs and to

prioritize conflicting goals. Since we adopt a unified

normative perspective both towards epistemic issues

and deontic issues, both being based on the notion of

norm as a default or defeasible rule, the axiological

component is conceptualized as a meta-level

Prioritized Default Logic (Brewka, 1994).

In a multi-agent environment the mutual update

of agents’ EDA models is essential as a result of

perceptual events, such as message exchange. There

is also the possibility of using shared spaces such as

the information fields mentioned in section 6, which

exist at an inter-subjective level. However, the

specification of the EDA update using a pragmatic

function is still the subject of current research, and

will be reported in the near future. A related line of

research that is being pursued at the moment

involves the software simulation of EDA models,

which raises some software engineering questions,

related to the implementation of heterogeneous

multi-agent systems implementation, where

interaction aspects become a key issue, requiring a

pragmatic interpretation of the exchanged messages.

REFERENCES

Belnap, N., 1991. Backwards and Forwards in the Modal

Logic of Agency. Philosophy and Phenomenological

Research, vol. 51.

Bratman, M., 1987. Intentions, Plans and Practical

Reasoning. Harvard University Press, USA.

Brewka, G., 1994. Reasoning about Priorities in Default

Logic. In Proceedings of AAAI-94, AAAI Press,

Seattle, USA.

Castelfranchi, C., 1993. Commitments: from Individual

Intentions to Groups and Organisations. In Working

Notes of AAAI’93 Workshop on AI and Theory of

Groups and Organisations: Conceptual and Empirical

Research, pp.35-41.

Conte, R., and C. Castelfranchi, 1995. Cognitive and

Social Action, UCL Press, London.

Habermas, J. 1984. The Theory of Communicative Action:

Reason and Rationalization of Society. Polity Press.

Cambridge.

Filipe, J., 2000. Normative Organisational Modelling

using Intelligent Multi-Agent Systems. Ph.D. thesis,

University of Staffordshire, UK.

Filipe, J. and K. Liu, 2000. The EDA Model: An

Organisational Semiotics Perspective To Norm-Based

Agent Design. Workshop on Norms and Institutions in

Multi-Agent Systems, Barcelona, Spain.

Filipe, J., 2002. A Normative and Intentional Agent Model

for Organisation Modelling. Proceedings of the 3rd

Workshop on Engineering Societies in the Agents

World, ESAW 2002. Madrid, Spain, também

publicado como post-proceedings em: Petta, Tolksdorf

and Zambonelli (Eds), Engineering Societies in the

Agents World III, Springer-Verlag, LNAI 2577, pp.

39-52, ISBN 3-540-14009-3.

Filipe, J., 2003. Information Fields in Organization

Modeling using an EDA Multi-Agent Architecture.

AAAI Spring Symposium 2003 on Agent-Mediated

Knowledge Management (AMKM-2003). Stanford,

USA.

Jennings, N., 1994. Cooperation in Industrial Multi-Agent

Systems, World Scientific Publishing, Singapore.

Liu, K., 2000. Semiotics Information Systems

Engineering. Cambridge University Press. Cambridge.

Liu, K., K. Nakata and C. Harty, 2010, Pervasive

informatics: theory, practice and future directions

Intelligent Buildings International, Volume 2, Number

1, 2010 , pp. 5-19(15).

Peirce, C. (1931-1935) Collected papers of Ch. S. Peirce,

C. Hartshorne and P. Weiss (Eds). Cambridge, Mass.

Rao, A. and M. Georgeff, 1991. Modeling Rational

Agents within a BDI architecture. In Proceedings of

the 2

International Conference on Principles of

Knowledge Representation and Reasoning (Nebel,

Rich and Swartout, Eds.),439-449. Morgan Kaufman,

San Mateo.

Singh, M., 1996. Multiagent Systems as Spheres of

Commitment. In Proceedings of the Int’l Conference

on Multiagent Systems (ICMAS) - Workshop on

Norms, Obligations and Conventions. Kyoto, Japan.

Stamper, R., 1973. Information in Business and

Administrative Systems. John Wiley & Sons.

Stamper, R., 1996. Signs, Information, Norms and

Systems. In Holmqvist et al. (Eds.), Signs of Work,

MULTI-AGENT SYSTEMS IN INTELLIGENT PERVASIVE SPACES

Semiosis and Information Processing in Organisations,

Walter de Gruyter, Berlin, New York.

Stamper, R., K. Liu, M. Hafkamp and Y. Ades, 2000.

Understanding the Roles of Signs and Norms in

Organisations – a Semiotic Approach to Information

Systems Design, Behavior and Information

Technolog, 19(1), 15-27.

Stamper, R. (2000). Information Systems as a Social

Science: An Alternative to the FRISCO Formalism. In

Information System Concepts - An Integrated

Discipline Emerging, E.D. Falkenberg, K. Lyttinen, A.

A. Verrijn-Stuart (eds). Proc. ISCO 4 Conference,

Kluwer Academic Publishers, pages 1-51.

von Wright, G., 1951. Deontic Logic. Mind, 60, pp.1-15.

von Wright, G., 1968. An Essay in Deontic Logic and the

General Theory of Action. Acta Philosophica Fennica,

21. North-Holland.

Wooldridge, M. and N. Jennings, 1995. Agent Theories,

Architectures and Languages: A Survey. In

Wooldridge and Jennings (Eds.) Intelligent Agents,

Springer-Verlag, LNAI 890.

Yu, E. and J. Mylopoulos, 1997. Modelling Organisational

Issues for Enterprise Integration. In Proceedings of

ICEIMT’97, 529-538.

ICISO 2010 - International Conference on Informatics and Semiotics in Organisations