MASTER WEB INTELLIGENCE Plan

1. Definitions
Multi-Agent Systems 2. Action Domains
3. Positioning
Introduction 4. “Vowels” Dimensions
5. Multi-Agent Engineering
Olivier Boissier 6. Perspectives …
Olivier.Boissier@emse.fr

© O. Boissier ENSM Saint-Etienne

1 © O. Boissier ENSM Saint-Etienne
2
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Multi-Agent Systems E-Alliance

Definitions Agent (in a Multi-Agent World) Definitions

Print-Shop Agent : physical or software, Environment

Alliance Goal: print
autonomous entity that is
pro-active, reactive,
reactive perception
Workshop Workshop Workshop social,
social able to take
U U U
C
part to an organised
C Contrat C1
N C A N A N A activity, in order to
Agent Platform
achieve its goals,
by interacting with action
A ‘‘Alliance’’ Agent C ‘‘Contract’’ Agent N ‘‘Negotiation’’ Agent U ‘‘User’’ Agent
Environment Interaction Organisation other agents and
Interaction
users
users.
Multi-Agent System (MAS) : set of agents, that interact with each
other, situated in a common environment, eventually, building or User
participating to, an organisation Organisation
© O. Boissier ENSM Saint-Etienne
3 © O. Boissier ENSM Saint-Etienne
4
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Autonomous Agent (in a Multi-Agent World) Definitions Multi-Agent Systems Principles Definitions

• An agent X is autonomous Environment • The Agent perspective (micro perspective)

with respect to Y for O in Goal: print
• Reactive & Pro-Active entities / Encapsulation of control
situation S perception
• Y can be a user, another agent,
• Autonomy: agents may exhibit activities that are not the one
a group of agents, an expected by the other agents in the system
organisation, …
• Delegation: agents may receive some control over their activities
• O can be a goal, a plan, an
action, a resource, a norm, a
role, …
• The Multi-Agent System perspective (macro perspective)
It means that: • Distribution of knowledge, of resources, of reasoning/decision
capabilities
• agent X can decide locally of the action
adoption of O in situation S • Decentralization (loose coupling) of control, authority
• And Y has no certainty that X is Interaction • Agreement technologies, Coordination models and mechanisms to
going to adopt O in situation S
install coordination between the autonomous agents
! Loose coupling between User
agents • Emergent / Social order / Normative functioning
Organisation
© O. Boissier ENSM Saint-Etienne
5 © O. Boissier ENSM Saint-Etienne
6
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Plan MAS Action domains Action domains

• Socio-technical Systems
1. Definitions • Integration of software applications, with humans, organizations and the
physical world
2. Action Domains • Making them interoperate, interact, cooperate in a flexible and consistent
manner with each other
3. Positioning
4. “Vowels” Dimensions • Problem Solving
Modeling and solving problems by cooperation between local solvers
5. Multi-Agent Engineering •
• Installing top-down and/or bottom-up (emergent) solving process
6. Perspectives …
• Simulation
• Modeling and reproducing complex phenomena of interacting entities in the
real world in order to understand or to explain their behavior

© O. Boissier ENSM Saint-Etienne

7 © O. Boissier ENSM Saint-Etienne
8
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Socio-Technical Systems (1) Socio-Technical Systems (2)
Action domains Action domains

• Industries, services, IT applications are getting global • Properties of the targeted applications:
• Placed at the centre of multiple networks • Absence of monolithic vision
• Developing Knowledge intensive processes • Incremental development, by different teams
• Based on large scale underlying IT platforms such as Internet, Web, Internet of • Multi-* (sites, expertise, domains, points of view, decisions, goals,
Things motivations, …)
• Continuous execution and adaptation
• Industries, services, IT applications are situated in an ever-evolving • User-Centred
environment
• Main requirements:
• Requiring efficient collaboration processes
• Openness, permeability, scalability in size or structure
• While keeping flexibility and agility
• Distribution, no central control, control and interaction are local
• Autonomous Interacting entities loosely coupled with others or applications
• Users are more and more at the centre of the cooperation and
• Knowledge Intensive processing and sharing
collaboration taking place in these socio-technical systems
• Users may delegate their decisions to the application

© O. Boissier ENSM Saint-Etienne

9 © O. Boissier ENSM Saint-Etienne
10
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Example (1/3) Action domains

Example (1/3) Action domains

Service Personnalisation

Plate Forme Agents (FIPA)

Agents
Médiateurs
Agents Utilisateurs
Agents Services Agent Agent
Utilisateur Profil

Agent Agent Agent

Source CLIMATE Industrial Workshop 26/4/99
Fournisseur Fournisseur Interface
Contenu Contenu
Italie Japon

Source CLIMATE Industrial Workshop 26/4/99

© O. Boissier ENSM Saint-Etienne

11 © O. Boissier ENSM Saint-Etienne
12
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Example (2/3) Action domains
Example (2/3) Action domains

User-Centred Service Service Composition Semantic Web

Coordination Semantic Services Agent Technology
Description
Secured Execution and
monitoring of services

Source : CASCOM Source : CASCOM

FP6-IST-2 FP6-IST-2

© O. Boissier ENSM Saint-Etienne

13 © O. Boissier ENSM Saint-Etienne
14
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Example (3/3) Action domains

MAS Action domains Action domains

Adaptation & optimisation • Socio-technical Systems

Planification, coordination, optimisation along a
bottom-up approach
• Integration of software applications, with humans, organizations and the
• Responsibility Delegation physical world
• Communication between the nodes • Making them interoperate, interact, cooperate in a flexible and consistent
• Real time detection & reaction to changes manner with each other
• Adaptation to changes & continuous optimisation

• Problem Solving
• Modeling and solving problems by cooperation between local solvers
• Installing top-down and/or bottom-up (emergent) solving process
Planification, coordination, optimisation along
a top-down approach :
• Centralised collect and processing of
• Simulation
informations and events • Modeling and reproducing complex phenomena of interacting entities in the
• Propagation of plans & decisions real world in order to understand or to explain their behavior
• No realtime decision.
Source Whitestein Agent Technology Conference 2004
© O. Boissier ENSM Saint-Etienne
15 © O. Boissier ENSM Saint-Etienne
16
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Problem Solving Action domains
Example (1/2) Action domains

• Properties of the targeted applications:

• Absence of global strategies, of global solving method
• Interaction between local strategies, between local solving
methods
• Solution is the result of the interaction between local processes
(points of view, decisions, goals, motivations, …)
• Continuous functioning and evolution

• Main requirements:
• Decentralisation, local control, interactions
• Openness, permeability, scalability in size or structure
Ferrand 97
• Shared and dynamic environment
• Emergence of the solution
© O. Boissier ENSM Saint-Etienne
17 © O. Boissier ENSM Saint-Etienne
18
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Example (1/2) Action domains

Example (2/2) Action domains

Design of Complex Systems

• Multi-Disciplinary Simulation & optimisation (ID4CS)
• Design of complex system :
• Multi-level, Multi-disciplinary
• Multi-methods
• Multi-objectives, Multi-attributes
• Uncertainty
• Cooperation methods between optimisation
technics,
i+1 i+2 • Management of uncertainty
i
i-1
i-2 • Multi-* problem solving
Ferrand 97 • Emergence

© O. Boissier ENSM Saint-Etienne

19 © O. Boissier ENSM Saint-Etienne
20
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 MAS Action domains Action domains
Example (1/2) Action domains

• Socio-technical Systems
• Integration of software applications, with humans, organizations and the
physical world In order to:
• Making them interoperate, interact, cooperate in a flexible and consistent
manner with each other

• Problem Solving
• Modeling and solving problems by cooperation between local solvers
• Installing top-down and/or bottom-up (emergent) solving process

• Simulation
• Modeling and reproducing complex phenomena of interacting entities in the Understand, Explain
real world in order to understand or to explain their behavior
Discover, …, Help,
© O. Boissier ENSM Saint-Etienne
21 © O. Boissier ENSM Saint-Etienne
22
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Example (1/2) Action domains

Example (2/2) Action domains

http://www.massivesoftware.com/
MANTA [Drogoul 93]
The Return of the King (2003)
The Two Towers (2002)
The Fellowship of the Ring (2001) I, Robot (2004)

Ratatouille
Organisation émergente (2007)

…, Entertainment
Conversational Zeno Robot
http://hansonrobotics.com/
© O. Boissier ENSM Saint-Etienne
23 © O. Boissier ENSM Saint-Etienne
24
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Plan History – Major Steps Positioning

• 1980 : Agents in the Artificial Intelligence (AI) area

1. Contexte •From AI to Distributed AI …
… to Multi-Agent Systems
2. Definitions
3. Action Domains • 1990 : Agents are invading other domains
• Personal Assistants, avatars,
4. Positioning • Mobile Agents,
5. “Vowels” Dimensions • Reactive Agents, ….
6. Multi-Agent Engineering
• 1995 : Agents spread in other domains, Application
7. Perspectives … domains are enlarging
• Artificial Life, Economic Agents, …,
25
• …, Web, Ambient Computing, … 26
Multi-Agent Systems: Introduction © O. Boissier ENSM Saint-Etienne Multi-Agent Systems: Introduction © O. Boissier ENSM Saint-Etienne

History – Evolutions Positioning

Inter-Disciplinary Domain… Positioning

• 1973 - 1980:
• Hearsay II (1973): blackboard architecture for speech recognition • Direct Links with:
• Actor Languages (1973): messages as control structures
• Beings (1975), Society of Minds (1978)
• Programming, Objects...
• Artificial Intelligence,
• 1980 - 1990:
• Contract Net (1980): hierarchical decentralized control
• Distributed Systems, Parallelism,
• DVMT (1984): Distributed Interpretation
• Subsumption architecture (1986) : Reactive Robots
• MACE (1987): multi-agent platforms • But also:
• 1990 - … :
• Complex System (physics, …, ethology, ecology, …)
• Self-organisation, emergence, Interactions, organisations, reputation, trust, • Artificial Life, Neural networks, …
Agent Oriented Software Engineering, …
• In 1995, first international conference ICMAS, • Social Psychology, Sociology, Activity Theory,
• since 2002, Autonomous Agents + MAS -> AAMAS Economy, ...
© O. Boissier ENSM Saint-Etienne
27 © O. Boissier ENSM Saint-Etienne
28
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Direct Inheritance Positioning
Multi-Agent vs Objects Positioning

• Object Oriented Programming: •An agent, as an object, encapsulates a state and

• Encapsulation, modularity : an object encapsulate data and behaviors
methods that manage them (ex : C++, Java, Smalltalk), BUT:
• Distribution : Distributed objects, CORBA, DCOM • An agent encapsulates its control over its behaviors; an object
• ! Actor Languages Development has only control over its state
• Interactions among agents have a broader scope than the
• Artificial Intelligence: method calls between objects. Interactions consist in goals,
• Symbolic Reasoning Models (Expert systems, Knowledge plans, actions, hypothesis exchanges
Representation), logic, … • An agent may have different control cycles (data-directed, goal-
• distribution : Blackboard Architectures directed, interaction-directed, …)
• A MAS has several control flows. An Object system has, a priori,
only one control flow.
• Distributed Systems
© O. Boissier ENSM Saint-Etienne
29 © O. Boissier ENSM Saint-Etienne
30
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Multi-Agent vs Artificial Intelligence (1) Multi-Agent vs Artificial Intelligence (2)

Positioning Positioning

Mono-agent perspective of Artificial Intelligence

Ex. dependence networks
is pushed away

goals:
➨ Knowledge, Goals, C

Actions gain a social actions : put_on goals: A actions : clear

dimension B

A
C
B Ag3

Ag1
Ag2

© O. Boissier ENSM Saint-Etienne

31 © O. Boissier ENSM Saint-Etienne
32
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Multi-Agent vs Artificial Intelligence (3) Multi-Agent vs Artificial Intelligence (4)
Positioning Positioning

Social Knowledge Social Interaction

Ag3 external description:
goals: on(A,B), actions: clear
resources: A, plans: on(A,B):=clear(C), Hello Ag1, I need your
put_on(A,B) action ‘‘put_on’’
Ag1 external description:
To set A on B, just do it!!!
goals: on(C,Table), actions: put_on
resources: B plans: on(C,Table):=clear(C)
...
A action dép. A
C put_on C
B Ag3 B Ag3 Command(Ag3,Ag1, put_on(A,B))
Ag1 Ag1
Ag2 Ag2

© O. Boissier ENSM Saint-Etienne

33 © O. Boissier ENSM Saint-Etienne
34
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Multi-Agent vs Distributed Systems Positioning A Large Domain!!! Positioning

• Both take into account interconnection and distribution From Autonomous Agents to Multi-Agent Systems
• Autonomous Robots
• In MAS, Interconnection and Distribution are concerned • Personal Assistants
by: • Desktop Agents
• The requirement of taking into account the agent autonomy, of
developing synchronization and coordination mechanisms to
• Softbots, Knowbots
coordinate their activities • Mobile Agents
• The requirement to represent and take into account the user • Reactive Agents
interests
• Intelligent Agents, Cooperative Agents, Conversational
• The requirement to cooperate and to achieve agreements (or Agents
even compete) with other systems aiming at achieving their own
interests. • Autonomous Agent in a multi-agent world

© O. Boissier ENSM Saint-Etienne

35 © O. Boissier ENSM Saint-Etienne
36
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Plan Multi-Agent Models

1. Definitions • Several multi-agent models exist in the literature !

2. Action Domains
• One possible structuration of the different models:
3. Positioning
4. “Vowels” Dimensions • Along four dimensions (‘‘Vowels’’ [Demazeau 95]) :
5. Multi-Agent Engineering • Agent, Environment, Interaction, Organisation

6. Perspectives …
• Taking two main points of view:
• global (System centred), local (Agent centred)

© O. Boissier ENSM Saint-Etienne

37 © O. Boissier ENSM Saint-Etienne
38
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Agent Model Global point of view Environment Model Global point of view

• To define the common space shared among the agents

• To define the autonomous entities in the system (internal • Required explicit abstraction so that agents can be part of a MAS (execution
architectures, knowledge representation, reasoning mechanisms…) space, coordination space, information space, etc)
• BDI Agents (Belief Desire Intention) [Rao 95] • Active entity that mediates the interaction among agents and provides access
• Rational Agents [Russel 91] to resources
• Situated Agents [Agre 87], [Maes 90]
• Sources : Simulation, Physics‡, …
• Sources : Artificial Intelligence, Objects, Robotics, ...
Agent Agent
Architectures Architectures

MAS MAS Environment

Services Services OSimulation/Services
O
Middleware Middleware

© O. Boissier ENSM Saint-Etienne

39 © O. Boissier ENSM Saint-Etienne
40
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Reference Model [Weyns 06] CARTAGO [Ricci 07]
Global point of view Global point of view

http://apice.unibo.it/xwiki/bin/view/CARTAGO/
• Perception &
Action Modalities,
• Resources,
Coordination
Artifacts
• Topology
• Proper Dynamic Artefact

© O. Boissier ENSM Saint-Etienne

41 © O. Boissier ENSM Saint-Etienne
42
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Interaction Model Global point of view Agent Communication Language GlobalNiveau

point of
Global
view

• To define and structure the dynamic relation between two or I call for
• High level language to share propositional BDI proposal
several agents through reciprocal actions that may draw patterns of attitudes, to collaborate, Negotiate, ...
activity (conversations) • FIPA-ACL [FIPA 02], KQML [Finin 97], etc. cfp
BDI
• Agent Communication Languages (ACL FIPA, KQML, …), Content Languages,
Ontologies I offer
• Set of performatives based on mental states
• Interaction Protocols, Conversations, … offer
• Inform, request, cfp, agree, understood
• Sources : Speech Acts, Conversations
Agent • Content Languages ACL
Architectures
• Ex: KIF, FIPA-SL, FIPA-CCL, etc.
• To express actions, objects, propositions Content Language
MAS Services Environment • Based on ontologies, i.e. common
Services O
O Interaction OSimulation/Services
O vocabularies relative to a domain (weather- ontology
Middleware ontology, cinema-ontology, etc)

Propose 10 NB
CFP 20 Couleur
© O. Boissier ENSM Saint-Etienne
43 © O. Boissier ENSM Saint-Etienne
44
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 ACL FIPA Message Example Interaction Protocol Example
Global point of view Global point of view

(inform communication act

:sender A used for routing the message
:receiver B used for routing the message
:content (price (bid goood02) 150) content
:in-reply-to round-4
:reply-with bid04
:encoding 1000
:language fipa-sl1 content language
:ontology hpl-auction
:reply-by 10 deadline for answering
:protocol offer interaction protocol
:conversation-id conv02 conversation id
)

© O. Boissier ENSM Saint-Etienne

45 © O. Boissier ENSM Saint-Etienne
46
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Organization Model Global point of view

Organization Modeling Language Global point of view

• To define the coordination pattern for achieving a given goal and To

Declarative definition of the agents organization expressing coordination patterns to
make these coordination patterns accessible to the agents
achieve a common global goal
• Organizational Structures : roles, groups, plans, scenes Ex : AGR [Ferber 98], Teamwork [Tambe 98], Islander [Esteva 01], Moise+ [Hubner 02], …
• To control and regulate the autonomous activities of the agents
Deontic Specification (DS)
• RBAC, Norms, Obligations, Permissions, Laws, …
• Sources : sociology, social psychology, CSCW,

Agent Agent
Architectures Architectures

MAS Services Environment

Services O
O Services O
O Interaction OSimulation/Services
O MAS
Services O
O Services O
O
Services
Interaction
Environment
OSimulation/Services
O
Organisation Organisation
Middleware Middleware

Structural Specification (SS) Functional Specification (FS)

© O. Boissier ENSM Saint-Etienne
47 © O. Boissier ENSM Saint-Etienne
48
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Organization Management Platform
Global point of view Synthesis Global point of view

Organization specification written with the organization modeling

language Common space shared among the agents
(Signals, Influences, Dynamics, …)
Management of the organizational entity
• Within the agents ex : Jason-MOISE [Hubner 07],
• Within components accessible to the agents
Autonomous entities
Environment
ex : Madkit [Gutknecht 00], Karma [Pynadath 03], Ameli [Esteva 04], S-Moise+ Interactions among
(Internal the agents (ACL,
[Hubner 05], SYNAI [Gateau 06], …
Architectures, Agent Interaction
http://moise.sourceforge.net Interaction
Knowledge, …) Protocols, …)
Organisation

Coordination Patterns (Organization Modeling Languages,

MAS Services O Services Environment Norms, …)
Services O Interaction OSimulation/Services
O
Organisation
Middleware

© O. Boissier ENSM Saint-Etienne

49 © O. Boissier ENSM Saint-Etienne
50
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Agent Model Local point of view

Agent types wrt Control Architecture
Local point of view

Agent : software or hardware encapsulating processing mechanisms and data,

which is able to control its decisions and actions (internal & external), to • Strength of the coupling of the decision mechanism
perceive and act on the environment, to interaction with the other agents and to with external events (environment, interactions,
manage and reason on the relations with other agents and norms.
organisation/normes)
goals
Situated Agents
• Reactive agent
•
• agents that reason on themselves and on the environment data

• Social Agents
• situated agents that reason also on the other agents

• Organization Aware Agents

•social agents that also reason on the organizations/norms in which
they are involved

© O. Boissier ENSM Saint-Etienne

51 © O. Boissier ENSM Saint-Etienne
52
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Reactive Agent Model Local point of view Reactive Agent Model (2) Local point of view

• Control Cycle: closed loop involving two operations "execute” and • Reactive approaches (developed in opposition to
"see" (Stimulus/Response) AI symbolic reasoning)
• reaction to the evolutions of the environment
• Different approaches based on :
• No explicit representation of the environment, of the other agents,
of its skills, …
• behaviours
• [Brooks 86], [Steels 89], (robotic)
• Decision taken using neither some past history nor future (no
• [Drogoul 93] (ethology)
planning)
• interactions
• [Demazeau 93] (image analysis, cartography, …)
see execute
Agent
• [Bura 91] (games)
• situations
• [Agre 87] (games)
Environnement • [Wavish 90] (design, manufacturing)
© O. Boissier ENSM Saint-Etienne
53 © O. Boissier ENSM Saint-Etienne
54
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Reactive Agent Model (3) Local point of view

Agent types wrt Control Architecture
Local point of view

• Strength of the coupling of the decision mechanism

Control Cycle Example (implemented using production rules)
with external events (environment, interactions,
condition-action rules organisation/normes)
goals
set of percepts
• Reactive agent
data
do {
percepts := see();
state := interpret_input(percepts);
rule := match(state,rules);
goals
execute(rule[action]); • Deliberative agent
} while (1); data

© O. Boissier ENSM Saint-Etienne

55 © O. Boissier ENSM Saint-Etienne
56
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Deliberative Agent Model Local point of view Deliberative Agent Model (2) Local point of view

• Introduction of “deliberate” function in the control cycle • Example of control cycle

between the "see” and "execute" functions in order to
s : state,
reason and choose the proper action
eq : event queue
• Explicit Representation of the environment, of the other
agents, of its skills, … s := initialize();
do {
• History management, planning options := option_generator(eq,s);
Agent deliberate selected := deliberate(options, s);
s := update_state(selected,s);
see execute execute(s);
state
eq := see();
} while(1);
Environnement

© O. Boissier ENSM Saint-Etienne

57 © O. Boissier ENSM Saint-Etienne
58
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Deliberative Agent Model (3) Local point of view

Agent types wrt Control Architecture
Local point of view

• BDI Control Cycle • Strength of the coupling of the decision mechanism

with external events (environment, interactions,
• b : beliefs, g : desires, i : intentions, eq : event queue organisation/normes)
goals
(b,g,i) := initialize(); • Reactive agent
data
repeat
options := option_generator(eq,b,g,i); goals
selected := deliberate(options, b,g,i); • Hybrid agent
i := selected ∪ i; data
execute(i); goals
eq := see(); • Deliberative agent
b := update_beliefs(b,eq); data
(g,i) := drop_successful_attitudes(b,g,i);
(g,i) := drop_impossible_attitudes(b,g,i);
forever
© O. Boissier ENSM Saint-Etienne
59 © O. Boissier ENSM Saint-Etienne
60
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Situated Agent Local point of view Social Agent Local point of view

• Action, • Interpretation of ACL, of interaction protocols

• Perception (ex: access to DataBase, internet, …, physical world, • Reasoning on interactions: interaction strategies Conv. With B
virtual world) • Conversation management
Conv. With C

Agent Agent
Architecture Architecture

Knowledge Perception Knowledge Perception

Interaction
CFP 20 Color
Beliefs Beliefs
Reasoning Desires, Goals Reasoning Desires, Goals
Intentions Intentions
Propose 10 BW

Action Action

© O. Boissier ENSM Saint-Etienne

61 © O. Boissier ENSM Saint-Etienne
62
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Organization Aware Agent Local point of view Synthesis Local point of view

• Reasoning on norms, on organizations [Carabelea 04] ACL interpretation

Skills
• Contracts, Dependence Networks [Sichman 94] Interaction protocols
Motivations
Conversation
• Reputation, Trust [Muller 06] Reasoning
Management
Knowledge Perception

Interaction
Agent Beliefs
Architecture Reasoning Desires, Goals
Intentions

A Organisation
Perception Action
Knowledge

D
Interaction

C Beliefs, Representation& Mngt of

A Reasoning Desires, Goals Event Perception Action on the
Organisations
Intentions
C B Representation & Mngt of environnement
B Organisation
Action Norms, Representation of the environment
D Social Reasoning
© O. Boissier ENSM Saint-Etienne
63 © O. Boissier ENSM Saint-Etienne
64
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Synthesis Synthesis
RS évalué KR ind. DS possibilités RS évalué KR ind. DS possibilités

évaluation raisonnement décision évaluation raisonnement décision

RS non évalué DS choix RS non évalué DS choix

CS ind. CS ind.

KR soc. engagement KR soc. KR soc. engagement KR soc.

Interprétation Exécution Interprétation Exécution

Interprétation CSsoc. Exécution Interprétation CSsoc. Exécution
Dialogue Dialogue Dialogue Dialogue
ACL, Pr ACL, Pr ACL, Pr ACL, Pr
PS ES PS ES
organisation organisation
KR soc. Org. émetteur effecteur KR soc. Org. émetteur effecteur
capteur récepteur capteur récepteur

Mécanismes de Représentation Connaissances Mécanismes de Représentation Connaissances

traitement traitement
Flux de données
© O. Boissier ENSM Saint-Etienne
65 Flux de données
© O. Boissier ENSM Saint-Etienne
66
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Synthesis Synthesis Local point of view

Compiled from AgentCore.java
public abstract class mast.core.AgentCore extends java.lang.Thread implements java.io.Serializable
{
public boolean RUNNING;
java.io.File log; ACL interpretation
java.lang.StringBuffer logString; Skills
long timeout;
java.lang.String name; Interaction protocols
mast.facet.Facet facet[]; Motivations
java.lang.Object facetParams[][];
java.util.Hashtable eventQueue; Conversation
java.util.Hashtable expectingAnswer;
java.util.Vector subscribeList; Reasoning
public mast.core.AgentCore();
public mast.core.AgentCore(java.lang.String); Management
public mast.core.AgentCore(java.lang.String,long);
public mast.core.AgentCore(java.lang.ThreadGroup,java.lang.String,long);
public void setFacets(mast.facet.Facet[], java.lang.Object[][]) throws Knowledge Perception
java.lang.IllegalThreadStateException;
public abstract void start(java.util.Hashtable, java.io.File);
Representation

Interaction
public void start(java.io.File);
public void run();
public void postEvent(mast.facet.FacetEvent); Beliefs
void forwardEvent(mast.facet.FacetEvent) throws mast.facet.InvalidEventException; & Mngt of Reasoning Desires, Goals
mast.facet.FacetDescription getNextReceiver(mast.facet.FacetEvent) throws
mast.facet.InvalidEventException; Event Perception
java.util.Vector getReceiversFor(mast.facet.FacetEvent); Organisations Intentions
boolean match(mast.facet.FacetDescription, mast.facet.FacetEvent);
public void setAnswerTo(mast.facet.FacetEvent, mast.facet.FacetEvent, boolean); Action on the
public void setNoAnswerTo(mast.facet.Facet, mast.facet.FacetEvent);
public void subscribe(mast.facet.Facet, mast.facet.FacetEventFilter); Representation Organisation
public void unsubscribe(mast.facet.Facet, mast.facet.FacetEventFilter);
public void unsubscribeAll(mast.facet.Facet); Action environnement
public void appendToLog(java.lang.String); & Mngt of
public void saveLogFile() throws java.io.IOException;
static {}; Representation of
} Norms,
Method mast.core.AgentCore() the environment
0 aload_0
1 aconst_null
2 ldc #1 <String "Agent Core">
Social
4 getstatic #2 <Field long DEFAULT_TIMEOUT>
7 invokespecial #3 <Method mast.core.AgentCore(java.lang.ThreadGroup,java.lang.String,long)>
10 return
Reasoning
Method mast.core.AgentCore(java.lang.String)
0 aload_0
1 aconst_null
2 aload_1

, , , ,
3 getstatic #2 <Field long DEFAULT_TIMEOUT>
6 invokespecial #3 <Method mast.core.AgentCore(java.lang.ThreadGroup,java.lang.String,long)>
9 return
Method mast.core.AgentCore(java.lang.String,long)
0 aload_0
X
1 aconst_null
2 aload_1
3 lload_2
4 invokespecial #3 <Method mast.core.AgentCore(java.lang.ThreadGroup,java.lang.String,long)>
7 return © O. Boissier ENSM Saint-Etienne
67 © O. Boissier ENSM Saint-Etienne
68
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Plan Multi-Agent Engineering
• Developing multi-agent applications is often a difficult task
implementation, distribution, communications, …
Definitions
•
1.
2. Action Domains • There exists
• Multiple technologies focused on particular points of a MAS
3. Positioning • Multiple agent programming languages, dedicated or general purpose based
on existing programming languages
4. “Vowels” Dimensions • Multiple multi-agent programming platforms, involving often a specific agent
architecture, proposing or not, first order abstractions for the environment, the
organisation, the interaction.
5. Multi-Agent Engineering • Multiple standards
• Multiple software engineering methods for the analysis and design of MAS.
6. Perspectives … ! Multiple languages, platforms, methods are available …
! But often limited to a very focused set of applicative domains.

! Which one to choose? How to choose? How to compare?

© O. Boissier ENSM Saint-Etienne

69 © O. Boissier ENSM Saint-Etienne
70
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Multi-Agent Technologies Technologies Declarative Approach Languages

• Agent Architectures and Theories • CLAIM (Computational Language for Autonomous

Intelligent and Mobile Agents)
• Coalition formation mechanisms • Cognitive Agent Programming Language
• Multi-Agent Planning • Belonging to Himalaya Framework (Hierarchical Intelligent
Mobile Agents for building Large-scale and Adaptive sYstems
• Agent Communication Languages, Interaction Protocols based on Ambients)
• Auction mechanisms • Based on process algebra in order to represent concurrency and
agent mobility
• Negotiation strategies and mechanisms, Argumentation • Based on SyMPA platform implemented in JAVA respecting the
MASIF standard
• Electronic Institutions, Organisations, Norms
• Reputation, Trust • FLUX :
• Mono & multi-agent Learning • Cognitive Agent Programming Language
• Fluent Calculus implementation (Action representation
• Self-organisation, emergence, … formalism)
• http://www.fluxagent.org
© O. Boissier ENSM Saint-Etienne
71 © O. Boissier ENSM Saint-Etienne
72
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Imperative Approach Languages Hybrid Approach Languages

• 3APL (An Abstract Agent Programming Language « triple-a-p-I »)

• JACK Agent Language (JAL) • Programming language for the development of cognitive agents:
• By defining structures for beliefs, goals, plans, actions (internal, external or communication)
• Developed by Agent Oriented Software and reasoning rules (modification of plan bases),
• By reasoning methods to generate plans, revise plans, to achieve goals
• Based on PRS, BDI model (similar to hybrid languages • Integration of Prolog and Java
such as Jason, 3APL, Jadex) • http://www.cs.uu.nl/3apl
• Jason : extended version of AgentSpeak(L) interpreter, agent oriented
• JAL is an extension of Java allowing to create plans, programming language based on logic. Introduced by Rao.
beliefs base, …
• Communication between agents based on Speech-act (beliefs and goals
• Possibility to use team of agents, organisation of annotated by the information sources)
agents • Plans annotations
• Functions for selecting, for trust computation,
• Functions and agent architecture may be adapted (perception, belief-revision, inter-
agent communication, acting)
• http://www.agent-software.com • Integration by the user of existing code by the way of internal
• Implemented in java, bound to the organisational language MOISE, interfaced with
the CARTAGO platform
• http://jason.sourceforge.net

© O. Boissier ENSM Saint-Etienne

73 © O. Boissier ENSM Saint-Etienne
74
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Existing Platforms Platforms

Standards Standards
• Platforms
• FIPA compliant
• Knowledge Sharing Effort The DARPA Knowledge Sharing
• FIPA-OS (http://sourceforge.net/projects/fipa-os/) Effort
• Jade/LEAP (http://jade.tilab.com/) • http://www-ksl.stanford.edu/knowledge-sharing/
• Others :
• SACI Simple Agent Communication Infrastructure (http://www.lti.pcs.usp.br/saci/)
• MASIF - OMG (Object Management Group) : OMG effort to
• Developping Environments
• Madkit (www.madkit.org) standardize mobile agents - middleware services and
• JADEX, BDI agent model based on JADE (http://sourceforge.net/projects/jadex) internal middleware interfaces
• JACK execution environment, compiler, BDI agent model based on • www.omg.org
Procedural Reasoning System (PRS) (http://www.agent-software.com)
• AgentBuilder based on Agent Oriented Program (AOP) (http://
www.agentbuilder.com/)
• AgentTool (http://macr.cis.ksu.edu/projects/agentTool/agentool.htm) • IEEE Computer Society FIPA Standards Committee
• ADELFE (http://www.irit.fr/ADELFE/)
(Foundation for Intelligent Physical Agents)
• Have a look at Software Products for MAS, AgentLink, June 2002 • www.fipa.org

© O. Boissier ENSM Saint-Etienne

75 © O. Boissier ENSM Saint-Etienne
76
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 FIPA Plateform JADE (Java Agent DEvelopment Framework)
Standards

Application Non agent software • Middleware for developing agent-based P2P application
Agent
• On fixed platforms, smart phones, …
Structure • Two main products :
Specification ACL
ACL Message Agent Agent Management
• Agent Platform compliant to
Specification
Library of FIPA specifications
communication
acts • API to develop agents in Java
Interaction ACL Agent Management Directory • Open Source Project, LGPL License
Protocols
+ System (AMS) Facilitator (DF)
Specification • Controlled by Telecom Italia Lab, who owns the project
CL
Content • Result of the joint effort of multiple actors belonging to the JADE Board (founded in
Language 2003) which missions concern the promotion, the governance and implementation
Specification SL HTTP
Message Transport Service IIOP of the changes of the JADE platform
Agent Message SMTP
Transport
etc.
Specification Plateforme Agent FIPA
• Project portal : http://jade.tilab.com
ACL = Agent Communication Language

© O. Boissier ENSM Saint-Etienne

77 © O. Boissier ENSM Saint-Etienne
78
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Standards & Multi-Agent Systems Standards Multi-Agent Methodologies Methods

• Ontologies : DAML, OIL, OWL, … The engineering of Multi-Agent Systems needs to take into account two levels:
Multi-Agent System level (System-Centred)
• http://www.daml.org •
• Number of agents, Agent Heterogeneity?
• http://www.ontoknowledge.org/oil/ • What is the common medium shared by the agents (Environment)?
• http://www.w3.org/ • What are the communication mechanisms between agents?
• What are the communication languages, the ontologies, the interaction protocols used by the
agents?
• Other standards (De Facto) • What is the organisation regulating the actions of the agents? How is it established?

• Jini (www.sun.com/jini), • How do the agents coordinate their actions? How to ensure a consistent behavior?
Agent level (Agent-Centred)
• UPnP (www.upnp.org), •
• What does an agent represent? What are the kinds of actions encapsulated into an agent?
• UDDI (www.uddi.org), • How do the agents represent the environment, the organisation in which they are situated?
• Salutation (www.salutation.org) • How do the agents process the interaction with other agents?

• mobility : Aglets (www.trl.ibm.com/aglets/) • What is the agent architecture?

• Web Services (http://www.w3.org/)

• …
© O. Boissier ENSM Saint-Etienne
82 © O. Boissier ENSM Saint-Etienne
83
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Tools supporting methods Methods Plan
• Software Engineering Tools supporting methodologies:
•
•
MASE AgentTool : macr.cis.ksu.edu/projects/agentTool/agentool.htm
ZEUS : sourceforge.net/projects/zeusagent
1. Definitions
•
•
Prometheus PDT : http://www.cs.rmit.edu.au/agents/pdt/
PASSI ToolKit : mozart.csai.unipa.it/passi/ptk.htm
2. Action Domains
• INGENIAS : grasia.fdi.ucm.es/ingenias/ 3. Positioning
• OPM : www.objectprocess.org
4. “Vowels” Dimensions
• Different ways to model applications:
• Agent Oriented Software Engineering 5. Multi-Agent Engineering
• Environment Oriented Software Engineering
• Interaction Oriented Software Engineering 6. Perspectives …
• Organization Oriented Software Engineering

© O. Boissier ENSM Saint-Etienne

84 © O. Boissier ENSM Saint-Etienne
85
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

To continue … Multi-Agent Modeling

• General references
• Pitfalls of Agent-Oriented Development, M. Wooldridge, N.R. Jennings, Agents ’98, 1998. • Multi-model :
• Foundations of Distributed Artificial Intelligence, G.M.P. Hoare, N.R. Jennings, Wiley & Sons, • Articulation of different formalisms
1996
• Les systèmes multi-agents, J. Ferber, InterEditions, 1995 • Multi-viewpoints :
• Multiagent Systems: A Modern Approach to Distributed Artificial Intelligence,
edited by Gerhard Weiss, MIT Press, 1999. ISBN 0-262-23203-0 • Extern/intern, system centred/agent centred
• Principes et architectures des Systèmes Multi-Agents, J.P. Briot, Y. Demazeau, IC2, Hermès, • Multiple views on a shared world
2001
• Some standards • Multi-levels
• Knowledge Sharing Effort http://www.cs.umbc.edu/kse/ • Via organisations, via the environment (MAS)
• OMG Agent Working Group http://www.objs.com/isig/agent.html
• FIPA http://www.fipa.org • Multi-scales
• W3C http://www.w3.org • temporal, spatial, …
• Some general adresses
• Collège SMA de l AFIA : http://sma.lip6.fr
• AgentLink : http://www.agentlink.org
• AgentCities : http://www.agentcities.org
© O. Boissier ENSM Saint-Etienne
86 © O. Boissier ENSM Saint-Etienne
87
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Scientific Challenges Applicative Challenges
Complexity “massive”
Applications MAS
“open”
MAS
“inter-Organisation-
nal”
MAS
Source : AgentLink Roadmap
“closed”
MAS
Short-term Middle-term Middle-term Long-term

• different goals • Learning of

• a “same” goal • coordination • generic coordination coordination
• adhoc coordination mechanisms mechanisms mechanisms
mechanisms • Multi-Agent Design • Standards for the • Emergent
• Ad-hoc design • Important number of design of MAS Coordination
• Scalability in si- agents, scalability • Scalability in
mulation domains, •Scalability
•Robustness,
•Openness
© O. Boissier ENSM Saint-Etienne
89 © O. Boissier ENSM Saint-Etienne
90
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Domain Overview (1/2) Domain Overview (2/2)

• International Conferences
– International Conference on Multi-Agent System (ICMAS) de 1995 à 2000, • Standards
– International Conference on Autonomous Agents and MultiAgent Systems
(AAMAS) depuis 2002. (http://www.aamas-conference.org/) • FIPA (Foundation for Intelligent Physical Agents)
• French Conferences (http://www.fipa.org/)
– Journées Francophones SMA (http://www.cerv.fr/jfsma08/) • Competitions
– Collège SMA de l’AFIA (http://sma.lip6.fr/)
• European Projects • http://www.robocup.org/
– AgentLink (réseau d’excellence www.agentlink.org), Roadmap (www.agentlink.org/
roadmap)
• http://www.rescuesystem.org/robocuprescue/
• Some ‘‘Success Stories’’
– Brahms (agentsolutions http://agentsolutions.com/home.htm) @ NASA Ames
Research Center
•
– Living Systems (Whitestein technologies http://www.whitestein.com) @ ABX
Logistics http://tac.eecs.umich.edu/association.html
– eSTAR (http://www.estar.org.uk/) intelligent robotic telescope network
– CalicoJack (http://www.calicojack.co.uk/)
– Review of Industrial Deployment of Multi-Agent Systems http://agents.felk.cvut.cz/ • http://www.lips.utexas.edu/art-testbed/
teaching/33ui2/on-aplications.pdf
© O. Boissier ENSM Saint-Etienne
91 © O. Boissier ENSM Saint-Etienne
92
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Panorama

Domain Overview (3/3) Bibliography

• Journals • [Agre 87] Agre, P. E. and D. Chapman (1987). " Pengi: An Implementation of a Theory of Activity" .
AAAI-87. The Sixth National Conference on Artificial Intelligence, Menlo Park, CA., Morgan Kaufman,
• Autonomous Agents and Multi-Agent Systems Los Altos, CA
• Artificial Intelligence
• [Bayardo 97] R.J. Bayardo et al.. InfoSleuth: Agent-Based semantic integration of information in open
• Knowledge Engineering Review and dynamic environments. Proc. ACM SIGMOD Intern. Conf. on Management of Data, 1997.
• International Journal of Agent-Oriented • [Brooks 86] R. Brooks, J.H. Connel: Asynchronous Distributed Control System for a mobile robot,
Software Engineering (IJAOSE) SPIE 727 Mobile Robots, 1986
• Web Intelligence and Agent Systems • [Bura 91] "Eco-Résolution: un Modèle de Résolution par Interactions" Bura S., Drogoul A., Ferber J.
An International Journal & Jacopin E. Actes de RFIA, Lyon, 1991.
• News • [Carabelea 04] Cosmin Carabelea, Olivier Boissier, Cristiano Castelfranchi: Using Social Power to
• Agent List Enable Agents to Reason About Being Part of a Group. ESAW 2004: 166-177
• http://www.cs.umbc.edu/agentslist/ • [Casare & Sichman, 05] S. J. Casare and J. S. Sichman, Towards a functional ontology of reputation,
Proceedings of AAMAS’05, 2005.
• Distributed Artificial Intelligence List
• DAI-List-Request@ece.sc.edu • [Castelfranchi & Falcone, 98] C. Castelfranchi and R. Falcone, Principles of trust for MAS: Cognitive
anatomy, social importance and quantification. Proc of ICMAS’98, pages 72-79, 1998.
• French list
• sma@loria.fr • [Chavez 96] K. Chavez, P. Maes, Kasbah: an agent marketplace for buying and selling goods.
Proceedings of PAAM'96, Practical Application
• http://sma.lip6.fr/
• [Conte & Paolucci, 02] R. Conte and M. Paolucci, Reputation in Artificial Societies. Social Beliefs for
Social Order, Kluwer Academic Publishers, G. Weiss (eds), Dordrecht, The Netherlands, 2002.
© O. Boissier ENSM Saint-Etienne
93 © O. Boissier ENSM Saint-Etienne
94
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Bibliography Bibliography
• [Fullam et al, 05] K. Fullam, T. Klos, G. Muller, J. Sabater-Mir, A. Schlosser, Z. Topol, S. Barber, J.
• [Demazeau 95] Y.Demazeau: From interactions to collective behaviour in agent-based systems. In Rosenschein, L. Vercouter and M. Voss, A Specification of the Agent Reputation and Trust (ART)
Proc. of the 1st European Conf. on Cognitive Science, Saint Malo, France, April, 1995, p. 117-132. Testbed: Experimentation and Competition for Trust in Agent Societies, Proceedings of AAMAS’05,
• [Demolombe & Lorini, 08] R. Demolombe and E. Lorini, Trust and norms in the context of computer 2005.
security: a logical formalization. Proc of DEON’08, LNAI, 1998. • [Gandon 02] Fabien Gandon, Laurent Berthelot, Rose Dieng-Kuntz, A Multi-Agent Platform for a
• [Durfee 97] Durfee, E. H., Kiskis, D. L., and Birmingham, W.P., "The Agent Architecture of the Corporate Semantic Web, AAMAS 2002, 6th International Conference on Autonomous Agents, 5th
University of Michigan Digital Library", IEE/British Computer Society Proceedings on Software International Conference on Multi-Agents Systems, 9th International Workshop on Agent Theories
Architectures and Languages, Eds Castelfranchi C., Johnson W.L., p. 1025-1032, July 15-19, 2002,
Engineering (Special Issue on Intelligent Agents) 144(1), February 1997.
Bologna, Italy.
• [Esteva 01] Marc Esteva, Julian A. Padget, Carles Sierra: Formalizing a Language for Institutions and
Norms. ATAL 2001: 348-366 • [Gateau 06] Benjamin Gâteau: Using a Normative Organisational Model to Specify and Manage an
Institution for Multi-agent Systems. EUMAS 2006
• [Esteva 04] Marc Esteva, Bruno Rosell, Juan A. Rodríguez-Aguilar, Josep Lluís Arcos: AMELI: An
Agent-Based Middleware for Electronic Institutions. AAMAS 2004: 236-243 • [Gutknecht 00] Olivier Gutknecht, Jacques Ferber: MadKit: a generic multi-agent platform. Agents
2000: 78-79
• [Fasli 04] : M. Fasli et M. Michalakopoulos, « e-Game: a generic auction platform supporting
customizable market games », IAT 2004. • [Herzig et al, 08] A. Herzig, E. Lorini, J. F. Hubner, J. Ben-Naim, C. Castelfranchi, R. Demolombe, D.
Longin and L. Vercouyter. Prolegomena for a logic of trust and reputation, submitted to Normas 08.
• [Ferber 98] Jacques Ferber, Olivier Gutknecht: A Meta-Model for the Analysis and Design of
Organizations in Multi-Agent Systems. ICMAS 1998: 128-135 • [Hubner 02] Jomi Fred Hübner, Jaime Simão Sichman, Olivier Boissier: A Model for the Structural,
Functional, and Deontic Specification of Organizations in Multiagent Systems. SBIA 2002: 118-128
• [Finin 97] KQML as an agent communication language Tim Finin, Yannis Labrou, and James
Mayfield, in Jeff Bradshaw (Ed.), ``Software Agents'', MIT Press, Cambridge, to appear, (1997) • [Hubner 05] Jomi Fred Hübner, Jaime Simão Sichman, Olivier Boissier: S-MOISE+: A Middleware for
Developing Organised Multi-agent Systems. AAMAS Workshops 2005: 64-78
• [FIPA 02] http://www.fipa.org/repository/aclspecs.html
• [Hubner 07] Jomi Fred Hübner, Jaime Simão Sichman, Olivier Boissier: Developing organised
multiagent systems using the MOISE. IJAOSE 1(3/4): 370-395 (2007)
© O. Boissier ENSM Saint-Etienne
95 © O. Boissier ENSM Saint-Etienne
96
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction
 Bibliography Bibliography
• [Klusch 00] M. Klusch: Information Agent Technology for the Internet: A Survey. Journal Data & • [Muller & Vercouter, 05] G. Muller and L. Vercouter, Decentralized Monitoring of Agent
Knowledge Engineering, Kluwer Academic, 36(3), 2000 Communication with a Reputation Model, Trusting Agents for trusting Electronic Societies, LNCS
• [Klusch 96] Klusch, M., Shehory, O., 1996, Coalition Formation Among Rational Information Agents, 3577, 2005.
1996, Proc. of MAAMAW, Eindhoven, LNAI Series Vol. 1038:204-217, Springer-Verlag • [Muller 08] G. Muller, L. Vercouter, Trust and Reputation, cours EASSS 08
• [Klusch 99] M. Klusch (ed.): Intelligent Information Agents. Springer, 1999 • [Nodine 99] M. Nodine and J. Fowler. An overview of active information gathering in Infosleuth. Proc.
Intern. Conference on Autonomous Agents, USA, 1999.
• [Klush 01] M. Klusch (ed.) Special issue on Intelligent Information Agents: Theory and Applications,
Intelligent Cooperative Information Systems, vol. 10(1&2), March 2001 • [Parsons 03] S. Parsons et P. McBurney, « Argumentation-based Communication between Agents »,
Communication in Multiagent Systems, M.-P. Huget (ed.), LNCS 2650, 2003.
• [Léger 99] Alain Léger, Marie-Pierre Gleizes, Hans Joachim Einsiedler. ABROSE : A Co-operative
Multi-Agent Based Framework for Electronic Marketplace. Dans : InfoWin, ACTS Project Infowin • [Pynadath 03] David V. Pynadath, Milind Tambe: An Automated Teamwork Infrastructure for
(AC113), Vol. -, p. 21-36, 1999. Heterogeneous Software Agents and Humans. Autonomous Agents and Multi-Agent Systems 7(1-2):
71-100 (2003)
• [Maes 04] P. Maes, Agents that reduce work and information overload, Communications of the ACM,
Vol. 37, no. 7, July 1994, special issue on Intelligent Agents. • [Rao 95] A. S. Rao and M. P. Georgeff, BDI-agents: from theory to practice, Proceedings of the First
Intl. Conference on Multiagent Systems, 1995
• [Maes 90] Maes, P. (1990). "Situated Agents Can have Goals." Designing Autonomous Agents .
Maes, P. (Ed.). Cambridge, MA., MIT Press: 49-70. • [Ricci 07] Alessandro Ricci, Mirko Viroli, Andrea Omicini. "Give Agents their Artifacts": The A&A
Approach for Engineering Working Environments in MAS. 6th International Joint Conference
• [McBurney 03] P. McBurney et S. Parsons, « Dialogue Game Protocols », Communication in
Multiagent Systems, M.-P. Huget (ed.), LNCS 2650, 2003. "Autonomous Agents & Multi-Agent Systems" (AAMAS 2007)

• [Mui et al., 02] L. Mui and M. Mohtashemi and A. Halberstadt, Notions of Reputation in Multi-agent • [Russel 91] Stuart Russell and Eric Wefald. Do The Right Thing. The MIT Press, Cambridge,
Systems: A Review, Proceedings of Autonomous Agents and Multi-Agent Systems (AAMAS'02), p. Massachusetts, 1991
280-287, 2002, C. Castelfranchi and W.L. Johnson (eds), Bologna, Italy, July, ACM Press, New York, • [Shardanand 95] U. Shardanand, P. Maes (1995): Social Information Filtering: Algorithms for
NY, United States of America. Automating ``Word of Mouth'', In: Proceedings of the CHI '95
© O. Boissier ENSM Saint-Etienne
97 © O. Boissier ENSM Saint-Etienne
98
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction

Bibliography Bibliography
• [Sichman 94] Jaime Simão Sichman, Rosaria Conte, Cristiano Castelfranchi, Yves Demazeau: A
Social Reasoning Mechanism Based On Dependence Networks. ECAI 1994: 188-192 • [Wurman 98] : P. Wurman, M. Wellman et W. Walsh, « The Michigan Internet AuctionBot: a
configurable auction server for human and software agents », Autonomous Agents 1998.
• [Smith 80] : R.G. Smith, « The contract net protocol: High-level communication and control in a
distributed problem solver », IEEE Transactions on Computers, C29 (12), 1980. • [Zeng 97] Zeng, D. and Sycara, K. (1997). Benefits of Learning in Negotiation, In: Proceedings of
AAAI-97
• [Sycara 00] K. Sycara, S. Widoff, M. Klusch, J. Lu: LARKS: Dynamic Matchmaking Among
Heterogeneous Software Agents in Cyberspace. Journal on Autonomous Agents and Multi- Agent • E-Game : http://csres43.essex.ac.uk:8080/elearn/eg/
Systems, Kluwer, 2000
• [Sycara 99] K. Sycara: In-context information management through adaptive collaboration of
intelligent agents. In M. Klusch (ed.), Intelligent Information Agents, Springer, 1999.
• [Tambe 98] Milind Tambe: Implementing Agent Teams in Dynamic Multiagent Environments. Applied
Artificial Intelligence 12(2-3): 189-210 (1998)
• [Terziyan 07] Vagan Terziyan, MIT Department, University of Jyvaskyla, AI Department, Kharkov
National University of Radioelectronics http://www.cs.jyu.fi/ai/vagan/index.html
• [Van Elst 01] Ludger van Elst and Andreas Abecker: Ontology-Related Services in Agent-Based
Distributed Information Infrastructures. In: Proceedings of the Thirteenth International Conference on
Software Engineering & Knowledge Engineering, June 13-15, 2001, Buenos Aires, Argentina, pp.
79-85.
• [Walton 07] Christopher Walton. Agency and the Semantic Web. Oxford University Press, 2007.
• [Weyns 06] Danny Weyns, Tom Holvoet: A Reference Architecture for Situated Multiagent Systems.
E4MAS 2006: 1-40
© O. Boissier ENSM Saint-Etienne
99 © O. Boissier ENSM Saint-Etienne
100
Multi-Agent Systems: Introduction Multi-Agent Systems: Introduction