Semantic Web Services Negotiation - PowerPoint Presentation

Document Sample
Semantic Web Services Negotiation - PowerPoint Presentation Powered By Docstoc
					Automated Negotiation
for Web Services

 Research Proficiency Examination
 Student: Paul Fodor
 Advisor: Professor Michael Kifer
 Stony Brook University
 Computer Science Department
                                    1
Overview
   Web Services
     Industry
             Technologies
     Research Directions—Semantic Web Services

   Introduction to Automated Negotiation
   Logics and Systems for Web Services
    Automated Negotiation
   Conclusion


                                                  2
Introduction
   Problems with current Web service
    contracts:
     syntactic descriptions
     limited automation support
   Goal:
     automation  of service negotiation by providing
      a framework that emulates human business
      negotiation practices for Web services


                                                    3
Automated Negotiation for Web
Services
   Web Services
     Industry
             Technologies
     Research Directions - Semantic Web Services

 Introduction to Automated Negotiation
 Logics and Systems for Web Services
  Automated Negotiation


                                                4
Web Services
   A collection of protocols and standards for
    exchanging data between Internet applications
   Advantages:
     Open, text-based, standard   technologies
     Inexpensive to implement
   Examples:
     Search engines (e.g., Google, Amazon)
     Maps (e.g., Google Maps, Yahoo Maps)
     Travel systems (e.g., Rent-a-Car + SouthwestAirlines)
     Web service broker sites (e.g., www.xmethods.net)


                                                         5
Automated Negotiation for Web
Services
   Web Services
     Industry
             Technologies
     Research Directions - Semantic Web Services

 Introduction to Automated Negotiation
 Logics and Systems for Web Services
  Automated Negotiation


                                                6
Industry Technologies for Web
Services
    Web Services Protocol Stack
    Service Discovery: centralizes services into a common registry
    where it publishes locations and descriptions of services (UDDI)
    Service Description: describes the public interface to a specific web
    service (WSDL)

    XML Messaging: encodes messages in a common XML (XML-RPC
    and SOAP)

    Service Transport: transports messages (HTTP, SMTP, FTP)



                                                                            7
Industry Technologies for Web
Services




                                8
Simple Object Access Protocol
(SOAP)
   Basic messaging
    framework
   Remote Procedure Call
    (RPC) pattern
   SOAP message has three
    parts
       envelope – wraps entire
        message
       header – optional additional
        information (e.g., security,
        routing)
       body – application-specific
        data being communicated


                                       9
Web Service Description Language
(WSDL)
   WSDL presents the methods a service
    provides and how they can be accessed
    A  collection of ports (communication
      endpoints)
     Each port has a port type with operations
      (message exchanges)




                                                  10
Universal Description, Discovery,
and Integration (UDDI)
   UDDI Business Registry (UBR)
     "white pages" – contact info, description
     "yellow pages" – classification information
     "green pages" – technical data

 Problem: lack of semantics
 Current Usage: dynamically bind client
  systems to implementations

                                                    11
Additional Technologies
 Web Services Security
 Web Services Policy Framework
 eXtensible Markup Language
 Business Process Execution Language for
  Web Services
 Choreography Description Language
 Service-oriented architecture


                                        12
Web Services Security
(WS-Security)
 Implements integrity and confidentiality
 Security protocols: SAML and Kerberos




                                             13
Web Services Policy Framework
(WSPolicy)
 Web Services metadata (policies)
 Usage:
     Service capabilities
     Service requirements (message preconditions)
     Security requirements




                                                14
eXtensible Markup Language
(ebXML)
   Family of XML based standards to provide an
    open infrastructure for electronic business
    information from a workflow perspective:
     ebXML Messaging
     ebXML Registry
     ebXML Business Process Specification Schema
     ebXML Collaboration Protocol Profile and Agreement




                                                           15
Business Process Execution Language
for Web Services (BPEL4WS)
   Programming in the large—the high-level state transition
    logic of a system
       when to wait for messages
       when to send messages
       when to compensate for failed non-ACID transactions
   WSFL (IBM) + XLang (Microsoft) =>
               BPEL4WS (submitted to OASIS)
   BPEL4WS: a specification of executable business
    processes and business process protocols in terms of
    their execution logic or control flow


                                                              16
Choreography Description Language
(CDL)
 Describes Web services interactions in
  terms of their externally observable
  behavior (i.e., message exchanges)
 Temporal ordering and logic dependence
  of messages
 Focuses on collaboration




                                           17
Service-oriented architecture (SOA)

 A new paradigm for Web services
  programming
 Describes business functionality as a set
  of shared, reusable services




                                              18
Automated Negotiation for Web
Services
   Web Services
     Industry
             Technologies
     Research Directions - Semantic Web Services

 Introduction to Automated Negotiation
 Logics and Systems for Web Services
  Automated Negotiation


                                               19
Semantic Web Services
   Semantic Web: gives meaning to Web content in
    a manner understandable by machines
   Web Services + Semantic Web =>
    Semantic Web Services
   Requirements:
     Standard specification
        ontology theory support

     Reasoning support
   Technologies: OWL, OWL-S, WSDL-S, WSMO,
    SWRL, RuleML, and SWSL

                                               20
Web Ontology Language (OWL)
   A markup language for publishing and sharing
    data using ontology on the Internet
   An extension of Resource Description
    Framework (RDF)
   Categories:
     OWL Lite—based on the logic SHIF(D)
     OWL DL—based on the Description logic SHOIN(D)
     OWL Full



                                                       21
Web Ontology Language for
Services (OWL-S)
   Four different ontologies:
     Service ontology
     ServiceProfile ontology
        defines non-functional properties of the service
        exposes the functionality by referencing Inputs, Preconditions
         and Results from the ServiceModel
     ServiceModel ontology
        describes the behaviour of the Service in terms of processes
         and control constructs
     ServiceGrounding ontology
        binds processes, inputs and outputs of the process model to
         a transport protocol described in a WSDL document


                                                                     22
Web Ontology Language for
Services (OWL-S)
   Resource                          ServiceProfile


                                            What it does?
                provides

                           Service
                                             How it works?
          How to access it?
 ServiceGrounding
                                      ServiceModel




                                                             23
OWL-S Stack




              24
WSDL-S
 Focus on creation of a framework for
  semantic Web Services and processes
  that is fully aligned with the current
  industry standards
 WSDL-S is an extension of WSDL with
  semantic enhancements



                                           25
  Web Services Modeling Ontology
  (WSMO)
     Web Service Modeling Framework (WSMF)

                      Objectives that a client may have
                      when consulting a Web Service

                                    Goals

Terminology                                                      Semantic description :
of the information                                               - Capability (functional)
used by all other    Ontologies                  Web Services    - Interfaces (usage)
components


                                   Mediators

                        Connectors between components
                        with mediation facilities for handling
                        heterogeneities                                             26
WSMO
   Web Service Modeling Language (WSML)
       Variants:
            WSML-Core—the intersection of Description Logic and Horn Logic
             (without function symbols and equality) extended with datatype
             support
            WSML-DL—extends WSML-Core to an expressive Description
             Logic SHIQ
            WSML-Flight—extends WSML-Core in the direction of Logic
             Programming
            WSML-Rule—extends WSML-Flight to include function symbols
            WSML-Full—unifies all WSML variants
       Syntax:
            An XML and an RDF syntax




                                                                          27
Semantic Web Services Language
(SWSL)
 A rule language and a process ontology
 SWSL-Rules is based on F-logic
     Supports   meta-reasoning (HiLog and
      reification extensions)
     Supports prioritized defaults and classical
      negation by incorporating Courteous Logic
      Programming


                                                    28
RuleML
 A rule language for the Web
 Modules:
     Datalogmodule
     The NAF (negation-as-failure) extension
      module
     The FOL (First-Order Logic)




                                                29
Semantic Web Rule Language
(SWRL)
 Proposal for a rule language for the
  Semantic Web on top of OWL-DL
 Syntactic restrictions
     Itis impossible to use predicates with arity
      higher than two
     Function symbols are not allowed




                                                     30
Automated Negotiation for Web
Services
   Web Services
     Industry
             Technologies
     Research Directions - Semantic Web Services

 Introduction to Automated Negotiation
 Logics and Systems for Web Services
  Automated Negotiation


                                               31
Introduction to Automated
Negotiation
   Negotiation - a process in which interested parties
    resolve disputes, agree upon courses of action, bargain
    for individual or collective advantages, and/or attempt to
    craft outcomes which serve their mutual interests
   Features:
       How many parties are involved?
       Are the negotiations private or public?
       Is third party intervention possible?
   Types:
       Advocacy negotiation (win-lose negotiation)
       Mutual gains bargaining (win-win negotiation)


                                                             32
Introduction to Automated
Negotiation
   Automated negotiation:
     ontology  - semantic meaning
     strategy - elaborate and systematic plan of
      actions
   Areas:
     Negotiation  Support Systems (NSS)
     Intelligent software agents


                                                    33
Negotiation Support System
(NSS)
   A software program to help human negotiators
    to make better decisions
   Helps each user formulate and evaluate an offer,
    and sends a counter-offer
   Features:
     initialproblem setup
     near-constant human input
     final decisions are left to human negotiators
   Example: Inspire (http://interneg.org/inspire)

                                                      34
Intelligent software agents
   Approaches:
     Focusing    on the intelligent agents
         teaching the agents strategies for negotiation
     Focusing
             on the environment rules and
     mechanisms




                                                           35
Intelligent software agents
—focusing on the intelligent agents
   The offers between two intelligent agents are updated
    during negotiations
   CMU sequential decision making task using Bayesian
    probability as the underlying learning mechanism
   Genetic algorithms
       Steps:
            Begin with a population of various, randomly generated negotiation
             strategies
            Employs these strategies against strategies of other agents in a
             round of bargaining
            At the end of a round the agent evaluates the performance of each
             strategy in its current population
            Crosses over strategies from the current population (parent
             population) to create a next generation of bargaining strategies
             (child population)

                                                                              36
Intelligent software agents
—focusing on the environment
 Insist on the protocol for offers and
  counter-offers
 Example—Kasbah (MIT marketplace)
     No machine learning
     Form-filled strategies on Internet




                                           37
Automated Negotiation for Web
Services
   Web Services
     Industry
             Technologies
     Research Directions - Semantic Web Services

 Introduction to Automated Negotiation
 Logics and Systems for Web Services
  Automated Negotiation


                                               38
Logics and Systems for Web
Services Automated Negotiation
   Formal languages to specify the behavior of
    each independent service provider and
    consumer
   Logics and languages
     Linear Logic (LL)
     Defeasible Logic and Courteous Logic Programs
     Concurrent Transaction Logic for Services (CTR-S)
     Peertrust language
   Automated negotiation for Web services with
    machine learning

                                                          39
Linear Logic (LL)
   Facility for representation of resources


   Matskin and Rao (2003) developed an architecture and a
    methodology for agent-based Web service discovery and
    automated composition with LL
     Web Services described with declarative specifications in OWL-S
    Γ;S⊢G
       Γ - LL axioms representing agent’s capabilities
       S - initial state
       G - the goal state of the agent



                                                                   40
LL —Example
   A seller agent S – sells two tickets for $100:
          GS = {Dollar100}
          SS = {Ticket2}
          ΓS = Φ
   A buyer agent B – buys two tickets with $100:
          GS = {Ticket2}
          SS = {Dollar100}
          ΓS = Φ
   Negotiation:
        S proposes an offer to B:
          (m1, S,B, Ticket2 ⊢ Dollar100)
        B generates an acceptance:
          (m2, S,B, Dollar100 ⊢ Ticket2)
        A central market place determines if the agents agree on the plan using partial
         deduction in LL (back and forward chaining steps)




                                                                                           41
Defeasible Logic and Courteous
Logic Programs
   Rule-based approach to reason with incomplete and
    inconsistent information
       strict rules
       defeasible rules - may not fire even when their premises are
        satisfied, because they are blocked by other rules
   Features:
       no disjunction in the rule head
       priorities on rules may be used to resolve conflicts among rules
       efficient implementation
       provides a formal basis for analysis




                                                                           42
Defeasible Logic and Courteous
Logic Programs—Example
   Rule1 : State(st), Counteroffer(c), Min_profit(mp),
    Utility(u), st = 2, c + mp < u/2
    → ACCEPT PROPOSAL
   Rule2 : State(st), Counteroffer(c), Min_profit(mp),
    Utility(u), st = 2, c + mp > u
    →∼ ACCEPT PROPOSAL
   Rule3 : State(st), Counteroffer(c), Min_profit(mp),
    Utility(u), st = 2, c + mp < 2u/3, Prelim_propose(bid),
    new_bid = (bid+c)/2
    → PROPOSE(new_bid)


                                                        43
Defeasible Logic and Courteous
Logic Programs—Systems
   Michigan Internet AuctionBot
     uses  Courteous Logic Programs to declare
      contracts, to organize auctions and to collect
      results
   SweetDeal
     incorporates
                 process knowledge descriptions
      with OWL ontologies for searching, selection,
      and composition of Web Services

                                                       44
Concurrent Transaction Logic for
Services (CTR-S)
   The intersection of two areas:
     process modeling
     automated contracting
   An extension of CTR
     incorporates game-theoretic concepts to represent
      the choice of action that can be made by a party other
      than the reasoner
   A contract is modeled as a workflow that
    represents the various possibilities for the
    service and the outside actors


                                                          45
CTR-S—Example
  buy ← pay_escrow ⊗ (finance | sell)

  deliver ← insured ∨ uninsured
  sell ← reserve_item ⊗ (deliver ∨ keep_escrow)
  finance ← (approve ⊗ (make_payment ∨ cancel))
        ⊓ (reject ⊗ cancel)




                                                  46
Peertrust Language
 For trust between the requester and the
  service provider
 Extends WSMO with trust information in
  the description of Semantic Web Services




                                         47
Peertrust language—Example
   A bookstore “Unibook” applies a discount only if the
    customer is a student at a recognized university:
                   discount(BookTitle)$Buyer ←
                      studentId(Buyer)@University@Buyer,
                      validUniversity(University),
                      studentId(Buyer)@University.
                   studentId(“Paul”)@“Stony Brook University”
                   studentId(“Paul”)$Requester ←
                      member(Requester)@“BetterBusinessBureau”@Requester.
       Several iterations:
            “UniBook” proves “Paul” that it belongs to the “BetterBusinessBureau”
            “Paul” knows enough to disclose his student card
            The negotiation succeeds, “Paul” gets the discount


                                                                              48
Automated Negotiation for Web
Services with Machine Learning
   Features
     The  structure of the negotiation algorithm has
      to be independent of the Web services
      functionality
     Use different negotiation algorithms within the
      same Web service
   Example: Bayesian equilibrium


                                                    49
Conclusion
   Entry point to emulate and automate human
    business negotiation practices

   Future direction: Semantic Web Services




                                              50

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:15
posted:7/30/2012
language:English
pages:50