Marc Esteva (IIIA)

1
WP3 Organisations
Marc Esteva (IIIA) Eva Onaindia (UPV) Sascha
Ossowski (URJC) Enric Plaza (IIIA) Juan Antonio
Rodríguez (IIIA)
www.agreement-technologies.org
2
WP3 Organisations

• Topics addressed
• Organisation-centric point of view How to design
  / learn, instantiate and evolve complex org.
  structures ?
• Agent-centric point of view How to act alone /
  as part of a team within complex org. structures?
  
• User-centric point of view How to assure the
  usability of complex org. structures?

3
WP3 Organisations

• Tasks
• T3.1 Autonomic Electronic Institutions
• T3.2 Group Planning
• T3.3 Deliberative Agreement
• T3.4 3D Electronic Institutions
• T3.5 Organisational Teamwork

• Management
• Identify interrelations between WP3 tasks
• so as to avoid duplication of work/ identify
  synergies

4
Task 3.1
Autonomic Electronic Institutions
Juan Antonio Rodríguez Aguilar (IIIA)
www.agreement-technologies.org
5
Autonomic Electronic Institutions
Overview

• Introduction Aims and scope
• Activities
• Planning

6
Introduction

• Research issue How to endow an agent
  organisation or institution with autonomic
  capabilities to yield a dynamical answer to
  changing circumstances through the adaptation of
  its norms.
• An autonomic computing system must configure
  and reconfigure itself under varying (and even
  unpredictable) conditions. System configuration
  or "setup" must occur automatically, as well as
  dynamic adjustments to that configuration to best
  handle changing environments.

7
Introduction

• Example Supply chain automation

8
Activities

• Negotiation models and strategies for
  self-configuration
• Flexible negotiation models to reach agreements
  on the fly to respond to new goals
• Connections
• Task 3.5 in WP3
• Input Techniques from WP2 and WP4
• Output eProcurement demonstrator in WP8
• Background Work at IIIA on supply chain
  formation
• Current team Jesús Cerquides, Jar

9
Activities

• Learning norm reconfiguration mechanisms in
  cooperative and competitive scenarios
• to agree on how to respond to changing
  circumstances
• Connections
• Task 3.5 in WP3
• Output eProcurement demonstrator in WP8
• Background Results of TIN project on Autonomic
  Electronic Institutions
• Current team
• N.Salazar, J.L.Arcos, Jar Evolutionary approach
• J.L.Fernández, J.L.Arcos Under construction
• M.Vinyals, J.Cerquides, Jar Under construction

10
Plan
 Activity M0 M6 M12 M18 M24 M30
             
Negotiation models for norm agreement            
             
Empirical evaluation (configuration models)            
             
Learning models for reconfiguration            
             
Empirical evaluation (reconfiguration models)            
11
Task 3.2
Agreement Planning
Eva Onaindia (UPV)
www.agreement-technologies.org
12
Task 3.2 Agreement planning
WP3 Organisations (Leader Sascha Ossowski
URJC) Task 3.1. Autonomic Electronic
Institutions.) Task 3.2 Agreement Planning
(DESIGN, ORGANIZATION, MODEL) Task 3.3.
Deliberative Agreement social choice and
collective judgment models for open MAS Task
3.4 3D Electronic Institutions (3DEI) Task 3.5
Mechanisms for Efficient Organisational Teamwork
WP4 Argumentation and negotiation (Leader
Lluís Godo IIIA) Task 4.1. Agreement
Logics Task 4.2. Real-Time agreements Task
4.3. CBR-based Mediating Agent Task 4.4 Planning
and scheduling capabilities for an agent
(SOLVING TECHNIQUES, COMPUTATION) Task 4.5
Agreement management with Data Mining
13
What is agreement planning?
Planning by negotiation or Negotiation by
planning?
1. Planning is the problem, negotiation is the
technique (new planning framework)2.
Negotiation is the problem, planning is the
method (traditional view of planning for solving
a particular problem)
14
Planning decisions agreements

• Which (actions different alternatives)
• When (temporal allocation of actions)
• How (resources)

• And also
• Negotiation cycle planning execution
• Time to reach the agreements (planning time)
• .

15
Where planning meets negotiation?
PLANNING NEGOTIATION
AGENTS Single-agent Multiple planning entities (humans, soft agents, ) Collaborative framework Distributed planning Multi-agent planning
KNOWLEDGE Perfect knowledge Global knowledge/Partial knowledge Sharable / Private / Hidden Distributed planning
GOALS Joint goals Joint goals ? collaborative work, joint negotiation for a common global purpose Independent goals ? collaborative work, individual negotiation for individual interests Single-agent goal ? collaborative work (??) Independent goals ? collaborative individual work (do ones own thing)
OPTIMIZATION Time, resources Negotiation elements
16
Task 3.2 Workplan

• Project report
• D3.2.1 Group planning agreements M36

• Our proposal Study the relation between planning
  and negotiation
• M8 Analysis and identification of components
• Agents, Knowledge, Goals, Optimization
• Survey
• M12 Discussion with negotiation people
• Planning needs introduced by negotiation
• Planning agreements between agents

17
Task 3.3
Deliberative Agreement
Enric Plaza (IIIA)
www.agreement-technologies.org
18
Task 3.3. Deliberative Agreement social choice
and collective judgment models for open MAS
Objectives. The overall goal of this task is to
analyze properties and develop mechanisms for
collective decision making in human and
artificial agents. The emphasis will be on tasks
that require complex agreements, i.e. involving
at least one of the following (i) argumentation
processes for deliberation, (ii) aggregation of
sets of interconnected judgments or (iii)
organizational division of labour.
Technologies. This goal requires the integration
of relevant contributions in social choice
theory, argumentation models, aggregation
procedures, organizational and institutional
models.
19
Task 3.3. Deliberative Agreement social choice
and collective judgment models for open MAS
Focus while social choice models typically focus
on aggregating individual preferences, we will
focus on judgment aggregation. Judgment
aggregation is new field that aims at finding
collective judgments on logically interconnected
propositions.
Year 1
Requirement Analysis Study of relevant aspects
in social choice theory, argumentation models,
aggregation procedures, and organizational and
institutional models.
Deliverable D3.3.1 Requirement analysis for
deliberative agreement. Month 12
Years 2-5
Aggregation procedures for interconnected judgments. Month 24
Deliberation and aggregation for interconnected judgments. Month 40
Institutional models for deliberative agreement Month 54
20
Task 3.3. Deliberative Agreement social choice
and collective judgment models for open MAS
Any group that attempts to manage a common
resource (e.g., aquifers, judicial systems,
pastures) for optimal sustainable production must
solve a set of problems in order to create
institutions for collective action.
As an example of application, we consider the
management of a common resource, namely the
aquifers of a region, and the joint decision is
whether or not a shortage of water for
consumption is likely (e.g. to implement a ruling
that changed the normal exploitation regime to a
restricted exploitation regime)
21
Task 3.3. Deliberative Agreement social choice
and collective judgment models for open MAS
Interconnected causal judgments.
22
Task 3.4
3D Electronic Institutions
Marc Esteva (IIIA)
www.agreement-technologies.org
23
T3.4 3D Electronic Institutions
24
T3.4 3D EI Example
25
T3.4 Goal

• Facilitate the integration (participation) of
  humans into MAS.
• Graphical visualization of norms and interaction
  context
• Humans and software agents collaboration.
• Extending the application domains of Virtual
  Worlds

26
T3.4 Goal

• Definition of a methodology for the construction
  of 3D Electronic Institutions
• specification of institutional rules
• design and development of the Virtual World
• Development of algorithms and software tools to
  support the methodology
• Autonomic 3D Electronic Institutions

27
T3.4 Run Time Architecture
28
T3.4 3D EI

• In collaboration with the University of
  Technology Sydney, University of Western Sydney
  and University of Barcelona.
• New PhD student Tomas Trescak (co-supervised
  with Inmaculada Rodriguez)

29
T3.4 WorkPlan

• Goal development of the execution environment
  for 3DEI.
• T1. Study of available Virtual Worlds Clients and
  selection of what to use in this project.
  Definition of 3DEI in the representation language
  of the chosen VW client. (Feb - May 2008).
• T2. Adaptation of the map generation algorithm
  (Jun - Jul 2008)
• T3. Implementation of the Causal Connection
  Server to connect AMELI with the chosen VW
  Client (Jul Oct 2008)
• T4 Deployment and testing of the system (Oct 2008
  Jan 2009)
• T5. Norm and Context Visualization (Nov 2008
  Jan 2009)

30
Task 3.5
Mechanisms for Efficient Organisational Teamwork
Sascha Ossowski (URJC)
www.agreement-technologies.org
31
Task 3.5 Characteristics

• Objectives (DoW)
• to study how organizational structures can
  improve and accelerate co-ordination processes in
  open environments.
• to study the effect of organizational regulation
  on the quality and flexibility of teamwork

• Activities (DoW)
• Design and implementation of micro-level
  mechanisms
• Design and implementation of macro-level
  mechanisms

• Deliverables (DoW)
• D3.5.1 Design and analysis of organisational
  structures. M18.
• D3.5.2 Micro-level mechanisms M24.
• D3.5.3 Macro-level mechanisms M40.
• D3.5.4 Implementation of mechanisms. M54.

32
Task 3.5 Previous work

• Organisational structures
• Decision making in teamwork
• Social dependence networks
• Mapping to bargaining theory
• Abstraction for design/specification
• Model MAS/DPS structures in terms of roles,
  interactions, etc.
• Specify dynamics in terms of role-playing
  relations etc.
• Similarity measure for trust models
• Determine confidence in similar roles etc. (e.g.
  for bootstrapping)
• Service descriptions in SOMAS
• Role-based matchmaking
• Service composition filters

33
Task 3.5 Research lines

• Challenges regarding organisational structures
• Ch1 Language Beyond roles and interaction
  hierarchies
• Ch2 Model Design vs. learning of
  organisational structures
• Ch3 Exploitation Beyond partner selection in
  two-sided interactions

• Activities
• A1 Study of complex org structure models (Ch1)
• A2 Learning (extensions of) org. ontologies
  (Ch2)
• A3 Implications for teamwork planning trust,
  filtering, (Ch3)
• A4 Org. structures for Probability Collectives
  (Ch3)
• A5 Effect of run-time org. information on the
  efficiency of teamwork (Ch3)
• A6 Adapt/extend simulation environments (Ch23)

34
Task 3.5 Workplan
 Activity M0 M6 M12 M18 M24 M30
             
A1 Study of complex org structure models            
A2 Learning org. ontologies              
A3 Implications for teamwork planning            
A4 Org. structures for Prob. Collectives            
A5 Effect of run-time org. information            
A6 Simulation environments              
           
35
Task 3.5 Relation to other WPs/Tasks

• Strong
• T3.2 Group Planning (Eva Onaindia)
• T6.1 Design of a MAS methodology based on org.
  concepts (C.A. Iglesias)
• WP5 Trust (Carles Sierra)
• T1.3 Scalable Methods for Semantic Service
  Coordination (Alberto Fernández)
• Medium
• T3.3 Deliberative Agreement (Enric Plaza).
• T3.1 Autonomic Electronic Institutions (Juan
  Antonio Rodriguez)
• T2.2 Individual Reasoning over normative systems
  (Pablo Noriega)
• T2.3 Declarative Specification of EIs (Marc
  Esteva)
• Other
• WP67 Tools Infrastructure

36
WP3 Organisations
Marc Esteva (IIIA) Eva Onaindia (UPV) Sascha
Ossowski (URJC) Enric Plaza (IIIA) Juan Antonio
Rodríguez (IIIA)
www.agreement-technologies.org