Docstoc

A Framework for Agent Collaboration in Multi-Agent Systems

Document Sample
A Framework for Agent Collaboration in Multi-Agent Systems Powered By Docstoc
					         A Framework for Agent Collaboration
               in Multi-Agent Systems
                            Submitted by:
              Mohamed Gamaleldin Atwany
                          Supervised by:
 Abdel-Aziz Khamis, Phd.             Magdy Aboul-Ela, Phd.
    Dept. of Computer and                   Dept. of Computer and
    Information Sciences,                   Information Systems,
     Cairo University                        Sadat Academy for
                                            Management Sciences

A thesis submitted to the Department of Computer Science, Institute
   of Statistical Studies and Research, Cairo University, in partial
fulfillment of the requirements for the degree of Master in Computer
                               Science

                             July 2002
    A Framework for Agent
  Collaboration in Multi-Agent
            Systems
        Mohamed.Atwany@acm.org
http://www.geocities.com/matwany/macf.ppt
The Agenda

1.   Introduction to Agents and Multi-Agent Systems

2.   Multi-Agent Collaboration

3.   Proposed Multi-Agent Collaboration Framework

4.   Proposed Framework Implementation

5.   The Case Study: e-Trade Agent Team

6.   Summary and Conclusion
Introduction to Agents and Multi-Agent Systems

Defining Agents
An agent is a virtual or physical computational entity that
   have partial representation of       possess skills and can offer
    the environment                       services

   perceive and act upon its            possess resources of its own
    environment                          driven by a set of tendencies
   may be able to reproduce itself      autonomous behavior
   can communicate directly with        Possess behavioral flexibility
    other agents                          and rationality
Introduction to Agents and Multi-Agent Systems

Types of Agents
   Cognitive Agents
     Intentional (Rational) agents
        Have explicit goals motivating their actions
     Module-based agents
        Reflexive cognitive agents
   Reactive agents
     Drive-based agents
       Directed by motivation mechanisms
     Agents
       Respond to stimuli from the environment, behavior guided by the
         local state of the world in which they are immersed
Introduction to Agents and Multi-Agent Systems

Defining Intelligent Agents
   Able to pursue its goals and executes its actions such that it
    optimizes some given performance measure
   Operates flexibly and rationally in a variety of environmental
    circumstances, given the information they have and their
    perceptual and effectual capabilities
   Has explicit goals motivating its action
Introduction to Agents and Multi-Agent Systems

OO Paradigm vs. Agent Paradigm
   Object is the basic unit             Agent is the basic unit
   Entity state definition is           Entity state defined via
    unconstrained                         Belief, commitments, goals
    Type of messages are                 Types of messages include
    unconstrained                         request, inform, query
   Abstraction level is lower           Abstraction level is higher
                                          and hence, it is more suited
                                          to the development of open
                                          systems
Introduction to Agents and Multi-Agent Systems

Defining Multi - Agent Systems
A multi-agent system is a system composed of number of interacting agents
  and characterized by being comprised of the following elements
   An environment

   A set of passive environment objects that agents can perceive, create, destroy and modify

   A number of agents representing system’s active entities

   A number of relations that link objects and agents to each other

   A number of operations that enables agents to perceive, produce, consume, transform and
    manipulate environment objects

   Laws of the universe
Introduction to Agents and Multi-Agent Systems

Key Issues in Multi - Agent Systems
   Communication
   Interaction
   Coordination interactions
   Cooperation interactions
   Negotiation interactions
   Organization interactions
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Communication
   A threefold problem involving knowledge of interaction protocol,
    communication language and transport protocol
   Forms the basis for interaction and social organization
   Speech Acts Theory
      views natural human language as actions (a suggestion, a commitment, or a reply)
      classified to types (Assertive acts, Directive acts, …etc.)
   KQML (content, communication, and message layers)
   Conversations
      Defined as a series of communications among different agents that follows a
       protocol and with some purpose
      A layered conversational model (protocol, conversation, and policy layers)
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Interaction
   An interaction situation is an assembly of behaviors
    resulting from the grouping of agents acting in order to
    attain their objectives, paying attention to the resources
    available to them and to their individual skills
   Occurs between two or more agents brought into a
    dynamic relationship through a set of reciprocal actions
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Coordination
   Refers to either
       a state of an agent community where agents’ actions fit well with each
        other or
       to the process of achieving a state of coordination within an agent
        community
   Agents coordinate their actions for four main reasons
       Agents require information and results other agents’ supply
       Limited resources have to be shared to optimize carried actions and
        try avoid possible conflicts
       Enables cost reduction by eliminating pointless actions and avoiding
        redundant actions
       Agents might have separate interdependent objectives that they need
        to achieve while profiting from goal interdependencies
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Cooperation
 Defined as coordination among non-
  antagonistic agents where participants
  succeed or fail together
 A cooperative situation is validated if either
     Adding  a new agent could result in an increase in
      performance levels of the group
     Agent actions serve to avoid or to solve potential
      or actual conflicts.
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Negotiation
   Defined as
     Interaction between agents based on communication for
      the purpose of coming to an agreement, or
     A process by which a joint decision is reached by two or
      more agents, each trying to reach an individual goal or
      objective, or
     Coordination among competitive or simply self-interested
      agents or,
     As a distributed communication-based search through a
      space of possible solutions.
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Negotiation
   Is much related to distributed conflict resolution and decision-
    making
   Requires agents to use a common language
   Supports cooperation and coordination between agents
   The Process:
       Agents make proposals
       Proposals are commented (refined, criticized, or refuted) by other
        agents
       other agents then communicate their possibly conflicting positions,
       Agents then trying to move towards agreement by making
        compromises or searching for alternatives
Introduction to Agents and Multi-Agent Systems
Key Issues in Multi-Agent Systems
 Organization
   Defined as an arrangement of relationships between
    components or individuals which produce a unit, or
    system, endowed with qualities not apprehended at the
    level of the components or individuals.
   An organization links, in an inter-relational manner,
    diverse elements or events or individuals, which
    thenceforth become the components of a whole.
   An organization ensures a relatively high degree of
    interdependence and reliability, thus providing the
    system with the possibility of lasting for a certain length
    of time, despite chance disruptions
Introduction to Agents and Multi-Agent Systems

Applications of Multi-Agent Systems
 Problem Solving
 Multi-Agent Simulation
 The Construction of synthetic worlds
 Collective robotics
 Kinetic program design
Introduction to Agents and Multi-Agent Systems

Collaboration in Multi-Agent
Systems
   Defined as forms of high-level cooperation that requires the
    (development of) mutual understanding and a shared view
    of the task being solved by several interacting entities
   Collaboration occur within a team of agents cooperating to
    achieve some collective goal.
   As a team of cooperating agents, participating agents
    succeed or fail together.
   Sharing a mental state within a team of agents enables
    reasoning about their beliefs, commitments, and intentions
    and hence, reason about the success or failure of
    collaboration.
Introduction to Agents and Multi-Agent Systems
Collaboration in Multi-Agent Systems
Multi-Agent Collaboration
Theories
   The Theory of Joint Intentions
      defines logic of rational action that is intended to be used as a
         specification of agent design
        The basic argument is that a joint activity is one that is performed
         by individuals sharing certain specific mental properties which
         affect and are affected by properties of the participants
   The Shared Plans Theory
      several deficiencies noted in Pollack’s mental state of plans
        Defines the concept of a shared plan
        Describes the entire web of a team’s intentions and beliefs when
         engaged in teamwork
   The Theory of Cooperative Problem Solving Process
        presents a model of cooperative problem solving (CPS)
        characterizes agents’ mental states leading them to solicit, and
         take part in, cooperative action
Introduction to Agents and Multi-Agent Systems
Collaboration in Multi-Agent Systems
 Multi-Agent Collaboration Frameworks
 GRATE
     a general framework that enables the construction of multi-agent
      systems for the domain of industrial process control
     Applications could be built very rapidly because much of the general
      domain behavior is already defined
 STEAM
     enables a team of agents to act coherently in a way that overcomes
      the uncertainties of complex, dynamic environments in which team
      members often encounter differing, incomplete and possibly
      inconsistent views of the world and mental state of other agents
 The Issue of Interoperability
     The frameworks does not support interoperability
 Open systems Readiness
     Heterogeneous agents, no pre-specified interaction protocols, no pre-
      specified organization
Introduction to Agents and Multi-Agent Systems
Collaboration in Multi-Agent Systems
 The Development of a Shared Mental State
The shared mental state consists of the following set of shared
  knowledge structures:
 a dependency graph of achievement goals
 a dependency graph of commitments to achieve these goals
 a dependency graph of actions believed to achieve these
  goals
 a dependency graph of commitments to these actions
 a dependency graph of intentions of actions agents are
  committed to achieve
 a dependency graph of mutual beliefs about goal relevance
  and achievement status, status of commitments, status of
  intentions, and status of actions
Proposed Framework
Proposed Framework for Multi-Agent Collaboration

 Scope
   Creating, sharing, and maintaining a shared mental state
    within a team of agents
 Objectives
   Framework based on a formal model of teamwork
   Support different phases of cooperative problem solving
   Transparent to existing interaction protocols and agent
    organizations
   Transparent to development environments
   Transparent to agent architectures
Proposed Framework

The Methodology
Is based on the observation that behavior can
 be analyzed without any knowledge of the
 implementation details
The proposed framework should be based on
 two teamwork models
The proposed framework should adopt a
 layered conversational model
   Proposed Framework

   Overal Object Model
                                                                 Conceptual Framework



                                                                                                                  Conversational Model

                                                                                           State Model
                            Framework Implementation

                                                                                               Query Interaction Protocol          Conversational Policy


                                                                                 Formal Teamwork Model           Conversational Pattern


     ACL                                                               Framework Implementation Derivation Mechanism


                              Extention Mechanism
        BRL
                                                                            MAS Systems
                                               XML Message Format

 Ontology
                                                                                                                    Cooperation Mechanism

Message Encoding & Decoding Facility                   Communication Mechanism
                                                                                                   Negotiation Mechanism

                                                                  Coordination Mechanism


                                                                                           Team Organization
Proposed Framework

Components
Define a pattern of interaction for information
 exchange that agents should follow
Define unambiguous rules for reasoning about
 agent and team behavior
Maintain a clear separation between the
 generic specification defined by the framework
 and possible implementations of that
 framework
Proposed Framework
Components
The State Model
              _

                                                     intend action
                                              drop intention to action

                                                       Execution           mutual belief that the goal is    End State
                                                                             achieved, unachievable, or
                                                                                        irrelevant

                                    drop intention
                                                                      jointly intend
                                    to main joint
                                                                       main action
                                        action
                                                      Post-Planning



                          joint                                                          drop joint
                     commitment to                                                     commitment to
                       main action           joint commitment to subaction               main action
                                          individual commitment to subaction

                                                         Planning




                          team                                                          no mutual

                       formation                                                        belief that

                       conditions                                                        team T
                                                 team cannot achieve P
                                                                                           can
                      are matched
                                                                                        achieve P
                                                       Pre-Planning
                                                                                                            State State
                                                                                       intiator agent
                                                                                   recognizes problem
Proposed Framework
Components
The Conversational Model Contents
 A Query Interaction Protocol
 A Set of Collaborative Conversational Patterns
 A Collaborative Conversational Policy
Proposed Framework
Components
The Conversational Model
Conversational Patterns
 Form Team Conversational Pattern

         This     conversational      pattern   defines   the        preconditions,     post -conditions,
         reasoning, and mental states used to form a team.

 Valid Team States


         Pre-Planning

 Pre-Conditions


          ·     the   team   formation   facilitator   agent   has    a   designated    goal to    achieve
                through team action.

 Step 1: Reasoning Performed by the Team Formation Facilitator Agent


          a.    belief that a candidate team is able to achieve designated goal.

          b.    attempt to solicit assistance in order to form team to achieve designated
                goal.

 Step 2: Messages Sent by the Team Formation Facilitator Agent


          c.    announce     to   candidate   team agents      its   attempt   to   solicit   assistance   in
                order to achieve designated goal.

 Step 3: Reasoning Performed by Other Candidate Agents


          d.    maintain belief about team formation facilitator announced attempt.
Proposed Framework
Components
The Conversational Model
Conversational Policy
The proposed agent conversation policy consists of the
  following components:
 Domain and problem specific rules to be defined by
  the agent developer.
 Teamwork rules explicitly defined by the proposed
  framework state model through the definition of
  possible team states and the rules for reasoning
  about team states.
 Teamwork rules defined by the Cooperating Problem
  Solving Theory
 Teamwork rules defined by the Joint Intentions
  Theory
Proposed Framework

Integration into MAS Architectures
  Scenario #1:                                          Agent X                            Agent X                            Agent X

  All agents use the same                               Internal                           Internal                           Internal
                                                    Representation A                   Representation B                   Representation B
  framework implementation
  Facilitators translate between
  internal representations and the
  framework BRL                       All Agents use                 Facilitator for                        Facilitator for
                                       Framework                       Internal                               Internal
                                     Implementation                Representation A                       Representation B
                                            E2



  Some agents use framework              Agent X                                 Agent Y                                      Agent Z
                                          Internal                                Internal                                     Internal
  implementation E1, while others    Representation A                        Representation B                             Representation B
                                        Framework                               Framework                                    Framework
  use framework implementation        Implementation                          Implementation                               Implementation
                                             E1                                      E1                                           E2
  E2.
  Facilitators translate messages
  between framework
  implementations E1 and E2

                                                          Facilitator for                                             Facilitator for
                                                         Implementation                                              Implementation
                                                            E1 to E2                                                   E2 to E1
Proposed Framework Implementation

The Components
 A behavior representation language for modeling agent
  mental behavior (BRL)
    Ontology
    Modal and temporal operators
    grammar
    The extension mechanism
 An agent communication language (ACL)
    Speech acts
    mechanisms
 A message interchange format
    XML based message format
    Mapping to BRL elements
 A set of facilitators and components
    XML message encoding/decoding facility
Proposed Framework Implementation
Components
Proposed Ontology – Object Model
Proposed Framework Implementation
Components
Proposed Ontology - Frames
Frame           Field Name          Description

Belief          belief id           unique identification

                agent id            unique identification for agent or team that owns this belief


Belief          Message clause id   unique identification   for   the message clause,   see next
                                    section


Goal            goal id             unique identification

                agent id            unique identification for agent or team that owns this belief

Commitment      commitment id       unique identification

                agent id            unique identification for agent or agent group that owns this
                                    commitment

                goal id or action   unique identification for goal (action) that an agent is
                id                  committed to achieve (execute)
Proposed Framework Implementation
Components
Modal and Temporal Operators
 Modal clause      Description

 achievable        Expresses the status of achievability of a given goal

 irrelevant        Expresses the status of relevance of a given goal

 Exist             Expresses the status of existence of a designated knowledge
                   element.

 Temporal clause   Description

 Added             Expresses the status of addition of a new designated knowledge
                   element.

 Dropped           Expresses the status of dropping a designated belief

 Attempted         Expresses the past status of attempt in the past of a designated
                   action as a result of an agent intention

 attempting        Expresses the current status of attempt in execution of a
                   designated action as a result of an agent intention

 achieved          Expresses the status of achievement of a given goal
Proposed Framework Implementation
Components
BRL Language Grammar

<language clause>    ::= <message clause> | <BDI clause>
<message clause>     ::= <added clause> | <exist clause> | <dropped
                          clause> | <irrelevant clause> | <achievable
                          clause> | <achieved clause> | <unknown clause>
                          | <attempted clause> | <attempting clause>
<exist clause>       ::= exist (belief clause id | goal clause id |
                          commitment   clause id | intention clause id |
                          attempt clause id, truth value)
<dropped   clause>   ::= dropped (belief clause id | goal   clause id |
                          commitment   clause id | intention clause id |
                          attempt clause id, truth value)
Proposed Framework Implementation
Components
Speech Acts and The Inquire Mechanism

          Initiator                  Participant A                Participant B




                      A. inquire




                                                     B.1 inform




                        B.2 inform
Proposed Framework Implementation
Components
A Structured Message format based on XML
<team_message team_id="trade_team" message_id="attempt_to_solicit_assistance_1">

    <meta_info content_type="framework-implementation-horn-clause"/>

    <delivery_info sender_id="buyer">

       <recipient_list >

           <recipient agent_id="merchant"/>

           <recipient agent_id="delivery"/>

       </recipient_list >

       <reply_to_list>

           <recipient agent_id="merchant"/>

           <recipient agent_id="delivery"/>

           <recipient agent_id="buyer"/>

       </reply_to_list>

    </delivery_info>

    <message_content>

        <message_clause
truth_value="true"
Proposed Framework Implementation
The Components
Message encoding/decoding facility object model
Proposed Case Study

Teamwork in e-Trade
 The Problem
 Trade as an Organization of Trade Agents
 Trade as an Interaction of Trade Agents
 Trade as a Task Environment
 Trade as a Cooperative System
 Trade as a Coordinated System
 Collaboration Within a Trade Team
Proposed Case Study
Teamwork in e-Trade
Mapping the Purchase Process to CPS Model Phases
Purchase Process Phase      CPS Model Phase          Description

Decision to buy          Recognition          The buyer recognize the
merchandise                                   potential for performing
                                              purchase

Negotiation phase        Pre-team             Buyer negotiates with a
                                              group of merchants to
                                              select the best offer, and
                                              then, construct a trade
                                              team in order to execute
                                              the transaction

                         Planning             Buyer and merchant agree
                                              on all purchase attributes

Exchange phase           execution            Trade team members
                                              execute purchase

Settlement phase                              transaction steps in a
                                              timely coordinated manner
Proposed Case Study

Teamwork in e-Trade
 A Knowledge-Level Model for Reasoning
  about Collaboration
    Consisting of a set of mental elements
     categorized into beliefs, goals, commitments,
     and intentions and their dependencies
    Specifying a number of inference rules that allow
     reasoning about teamwork state
    Enable exchange of beliefs, goals, intentions,
     and commitments
Proposed Case Study

Teamwork in e-Trade
Trade Team Goal Hierarchy
                                               perform trade
          goal dependency
                                                 Goal G1




 perform payment                                                                             settle buyer part of
       G1.1                                                                                       transaction
                                                                                                     G1.6

                       receive
                      payment
                        G1.2
                                   deliver
                                 merchandise                       settle merchant part of
                                    G1.3               receive            transaction
                                                     merchandise             G1.5
                                                        G1.4
Proposed Case Study

Teamwork in e-Trade
A Teamwork Knowledge-Level Model
Agent Goal Attributes
 Goal Identification
 Agent Identification
 Goal Type
 Goal Addition Trigger
 Goal Drop Trigger
Proposed Case Study

Teamwork in e-Trade
A Teamwork Knowledge-Level Model
Agent Actions Attributes
 Actions Identification
 Agent Identification
 Parent Action
 Actions Type
 Actions Dependency
Proposed Case Study

Teamwork in e-Trade
A Teamwork Knowledge-Level Model
Agent Commitment Attributes
 Commitment Identification
 Agent Identification
 Commitment Type
 Commitment Addition Trigger
 Commitment Drop Trigger
Proposed Case Study

Teamwork in e-Trade
A Teamwork Knowledge-Level Model
Agent Intention Attributes
 Intention Identification
 Agent Identification
 Intention Type
 Intention Preconditions
 Intention Post ConditionsA
Proposed Case Study

Teamwork in e-Trade
A Teamwork Knowledge-Level Model
Agent Belief Attributes
 Belief Identification
 Agent Identification
 Belief Addition Trigger
 Belief Drop Trigger
Proposed Case Study

Teamwork in e-Trade
 An Approach for the Specification and
  Development of MAS Collaborative Behavior
               Develop prototype                                     Develop Input Scenarios




                                                 Verify prototype




                                   Develop MAS




                                                                Verify MAS
Proposed Case Study

Teamwork in e-Trade
 Collaborative Analysis Facility
    Based on proposed framework and proposed
     framework implementation
    Encode agent strategy with the prototype
    All possible collaborative scenarios are
     encoded into program input
    Validate generated behavior
Proposed Case Study

Teamwork in e-Trade
 Case Study Results
      Agent interaction and communication is crucial for
       maintaining a shared and consistent view of the trade
       problem
      A common view of the goals, actions, commitments, and
       intentions help agents reason on teamwork activities and
       state
      The use of conversational model helped agents reason about
       teamwork activities and state
      The implementation enabled agents to express collaborative
       mental behavior, using a set of agent interaction
       mechanisms, and transmitted using a common message
       format
      By reviewing generated output, the MAS developer is able to
       verify team and individual collaborative behavior
Summary and Conclusion
   Multi-agent systems are complex systems consisting of a
    number of agents, each of which by itself might represent an
    organization consisting of one or more agents
   within a MAS, agents interact in order to achieve their
    individual and collective goals in an act defined as
    collaboration
   The lack of support for interoperability and open systems in
    existing environments
   Proposed framework provides transparency to current
    development environment, interaction protocols, and agent
    organizations
Summary and Conclusion
   Separating framework from possible implementation enables
    the evolution of implementations that match environment
    needs
   The proposed implementation provides an ACL, a BRL, a
    message format, and a message encoding/decoding facility
   The agent paradigm suitable for the developing of open
    systems as found in the domain of e-Trade
   The proposed framework and the proposed framework
    implementation enabled the development of a MAS in the
    domain of e-Trade
   The proposed iterative approach eases the process of
    specification and development of MAS collaborative behavior
Questions
Thank You

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:2
posted:4/10/2012
language:
pages:54