Semantics to energize                  the full Services Spectrum:Ontological approach to better exploit services at techn...
Shift in Labor from Agriculture and Mfgto Service in Major Economies
Perspectives on Measurement of WorkService systems, service scientists, SSME, and innovationP. Maglio, S. Srinivasan, J. K...
Challenges • Business/Organizational Challenges“Each enterprise will measure and aspire to its own uniquelevel of dynamism...
Semantic Services Sciences (3S Model)• Based on IBM’s vision [1] of service  sciences  – Need to take a pervasive view of ...
Using the 3S Model• Consider global IT service provider developing a  new multimedia service for UK telecom provider  – Si...
Semantic Services Sciences (3S Model)
Ontologies to Describe Service Semantics                              (ontologies are about agreements)                   ...
Outline• Semantics for Technical Services  –   Data Semantics *  –   Functional Semantics *  –   Non Functional Semantics ...
Semantics for Technical Services     Current and past focus of METEOR-S
Semantics for Technical Services • Data/Information Semantics   –   What: (Semi-)Formal definition of data in input and ou...
Semantics for Technical Services                   Execution,      Development                  Adaptation and   / Descrip...
Semantics for Technical Services                  Execution,                      Development                 Adaptation a...
Semantics for Technical Services                  Execution,                   Development                 Adaptation and ...
Semantics for Technical Services                  Execution,                   Development                 Adaptation and ...
Semantics for Technical Services                  Execution,                  Development                 Adaptation and  ...
Semantics for Technical Services                  Execution,                                    Development               ...
DATA SEMANTICS
Data Semantics                           UDDI Query                                                        UDDI           ...
Data Semantics - options• Pre-defined agreement on all data fields  – Limited flexibility, hard to integrate new suppliers...
WSDL-S Specification(Now the key input to W3C leading to  Semantic Annotation of WSDL-             SAWSDL)
PurchaseOrder.wsdls…………<xs:element name= " OrderConfirmation" type="xs:stringwssem:modelReference=" rosetta#PurchaseOrderR...
Representing mappings <complexType name="POAddress"wssem:schemaMapping=”http://www.ibm.com/schemaMapping/POAdd            ...
FUNCTIONAL SEMANTICS
Functional Semantics                          UDDI Query                                                        UDDI      ...
Functional Semantics• Keyword based search in UDDI   – Needs human involvement   – Low precision and high recall• Port Typ...
Semantic Templates                                        Part of Rosetta                                                 ...
Semantic Discovery    • Finds actual services matching semantic templates    • Implemented as a layer over UDDI [1]    • C...
Non Functional Semantics  Business and Application constraints
Non Functional Semantics                           UDDI Query                                                        UDDI ...
Non Functional Semantics• Does the supplier support customer’s business  constraints   – e.g. cost, supply time etc.• Inte...
Non Functional Semantics•    Used in lifecycle    – Agreement Matching      •   Matching syntactically heterogeneous by se...
SWAPS: Use of Semantics in Agreement Matching An agreement is a collection of alternatives. A={Alt1, Alt2, …, AltN} An alt...
WS-Agreement Definition and Ontology  hasGuaranteeTerm                     GuaranteeTerm          An agreement consists of...
SWAPS Ontologies  WS-Agreement: individual agreements are   instances of the WS-Agreement ontology  Temporal Concepts: t...
Using Semantic Agreements withWSDL-S                              WS-Agreement Ontology                              Agric...
EvaluationConsumer       Provider        Approach 1:   Approach 2:     Approach 3:         Approach 4:Requirement    Capab...
The Matching ProcessObligated: Provider                                  Obligated: Provider99% of responseTimes <        ...
The Matching ProcessObligated: Provider                                     Obligated: ProviderresponseTimes < 14 s       ...
The Matching Process                                  isEquivalentObligated: Provider                               Obliga...
The Matching ProcessObligated: Provider                                     Obligated: ProviderresponseTime < 14 s        ...
The Matching ProcessObligated: ProviderresponseTime < 14 s                      Consumer                            Obliga...
The Matching ProcessObligated: ProviderresponseTime < 14 s                      Consumer                            Obliga...
The Matching ProcessObligated: ProviderresponseTime < 14 s                      Consumer                            Obliga...
The Matching ProcessObligated: Provider                                                                   notSuitable     ...
The Matching ProcessObligated: Provider                               Obligated: ProviderresponseTime < 14 s              ...
Dynamic Process Configuration• Operations Research has been used in industry  for business process optimization• There is ...
Dynamic Process Configuration Find optimal partners for the process based on process constraints – cost, supply time, etc....
Dynamic Process Configuration  Research Challenges          – Capturing functional and non-functional            requireme...
Abstract Process Specification                 1. Specify process control                    flow by using virtual        ...
Process Constraints• Constraints can be specified on a partner,  an activity or the process as a whole.• An objective func...
Process Constraints        Feature               Scope         Goal        Value      Unit         AggregationCost (Quanti...
Constraint Analysis• Multi-paradigm proposed:   – Integer Linear Programming for quantitative constraints   – Semantic Web...
Domain Ontology – Detailed View
Rules•    Supplier 1 should be a preferred supplier.    –   “if S1 is a supplier and its supplier status is preferred then...
Configuration Step 1: SemanticDiscovery
Configuration Step 2: QuantitativeConstraint Analysis
Configuration Step 3: Logical ConstraintAnalysis
EXECUTION SEMANTICS
Execution Semantics                            UDDI Query                                                         UDDI    ...
Process Adaptation• Ability to adapt the processes from failures,  unexpected events• Two kinds of failures      – Failure...
Process AdaptationAdaptation Problem                                                Optimally react to events like delays ...
Marginalizing events
Generating States using preconditionsand effects   Actions                               Chance Variables                 ...
Generated State Transition DiagramState   Values of Boolean             Explanation No.        variables1                 ...
Costs and Probabilities• Costs of ordering taken from configuration  module  – From first two service sets     • Optimal s...
Supplier Policy  – The supplier gives a probability of 55% for delivering the    goods on time.  – The manufacturer can ca...
Costs and Probabilities Current State      Action        Next State         Cost<O C R Del Rec >     NOP     <O C R Del Re...
Handling Coordination Constraints• Since the RAM and Motherboard must be  compatible, the actions of service managers  (SM...
Centralized Approach• State space created by Cartesian  product of transition diagrams• Joint actions from each state• Tra...
Decentralized Approach• Simple coordination  mechanism• If one service manager  changes suppliers  – All dependent manager...
Hybrid Approach• If the policy of some SM dictates it to change suppliers, the  following actions happen:   –   it sends a...
Evaluating Process Adaptation• Evaluation with the help of the supply chain  scenario• Two main parameters used for the ev...
Evaluating Adaptation
Semantics for Lightweight Services
Lightweight services and Mashups• REST based implementation becoming popular   – SOAP -> Web service   – REST -> Lightweig...
Current limitations and Role ofsemantics• Current Mashups tightly coupled (lack dynamism)   – E.g. HousingMaps.com uses cr...
An example• Consider a mashup: mybook.com  – Allows users to search and buy used and new books  – Gets data from various v...
An Example of Smashup (Semantic mashup)
Semantics for Knowledge Services     Current and past focus of METEOR-S
Semantics for Knowledge Services• Work in last two decades on knowledge  modeling not so successful  – Focus on capturing ...
High Level Model for KnowledgeServices
Using Model for Knowledge Services• Such a model can be used to answer  questions  – Find managers who have led project wo...
Autonomic Web Processes• The goal (Albatross)  – Self Configuring, Self Healing, Self Optimizing,    Self Protecting Busin...
Conclusions• Businesses perceive IT as an extension of  business strategy  – 3S Model uses semantics to provide a    compr...
Semantics to energize  the full Services Spectrum: Ontological approach to better exploit services at technical and busine...
Upcoming SlideShare
Loading in …5
×

Semantics to energize the full Services Spectrum: Ontological approach to better exploit services at technical and business levels

485 views

Published on

Amit Sheth's Talk at the inauguration invitational event for Research School on Services-Oriented Systems Engineering, Hasso-Plattner-Institute, Potsdam, Germany, June 22, 2006.

Published in: Education
  • Be the first to comment

Semantics to energize the full Services Spectrum: Ontological approach to better exploit services at technical and business levels

  1. 1. Semantics to energize the full Services Spectrum:Ontological approach to better exploit services at technical and business levels Amit Sheth LSDIS Lab, University of Georgia, Athens, Georgia, UGA Special Thanks: Kunal Verma
  2. 2. Shift in Labor from Agriculture and Mfgto Service in Major Economies
  3. 3. Perspectives on Measurement of WorkService systems, service scientists, SSME, and innovationP. Maglio, S. Srinivasan, J. Kreulen, J. SpohrerCACM July 2006.
  4. 4. Challenges • Business/Organizational Challenges“Each enterprise will measure and aspire to its own uniquelevel of dynamism based on its individual purpose. It is – How to effectively create new businessabout solutions using a global workforce being nimble and adaptable. A fully integratedbusiness platform canIT more faster, and completely, to – How to make respond responsive to businesschange. Whether it involves fulfilling a new mandate or strategyembracing a new market opportunity. Some organizations • Technical/Tactical Challengeswill push the envelope, automating event-triggered – How for highly integrated closed-loop processes,responses to add more dynamism in business process creationsetting the stage for self-optimizing systems.” – How to make processes adapt with changing environmentsSandra Rogers, White Paper: Business Forces Driving Adoption of Service OrientedArchitecture, Sponsored by: SAP AG
  5. 5. Semantic Services Sciences (3S Model)• Based on IBM’s vision [1] of service sciences – Need to take a pervasive view of services – Modeling people and organizational aspects as well as technical aspects of services• The 3S model [2] – Semantics for all types of services: Technical/Web Services to Knowledge Services [1] IBM, Services Sciences, Management and Engineering, http://www.research.ibm.com/ssme/ [2] Amit P. Sheth, Kunal Verma, Karthik Gomadam, Semantics to energize the full Services Spectrum, Communications of the ACM (CACM) special issue on Services Science, July 2006
  6. 6. Using the 3S Model• Consider global IT service provider developing a new multimedia service for UK telecom provider – Similar service already successfully provided in Japan• To provide the new multimedia service – Business manager must leverage assets – Human assets • Teams in China (Telco Equipment), India (Telco SW, Back Office) • People who have domain expertise in the new market • Project Management, … – Technical assets • Reuse SW assets and compose services to create technical platform • Use lightweight services for information aggregation and GUIs
  7. 7. Semantic Services Sciences (3S Model)
  8. 8. Ontologies to Describe Service Semantics (ontologies are about agreements) Autonomic Web Process*Organization Aspect of Agreement Strategy Layer Strategy Layer (Corporate Strategy and • Self Healing Goals) Requirement: Only Provide customer • Agile Operational Layer (Modeling Business support to gold customer Process to provide business services)People • Self Optimizing IT Layer Execution Layer (SOA Based IT Processes Requirement: and Services) S R- If cost > $$$$, • Self Configuring O TE t ME ou customer = gold Implementation Layer (Databases, OS, etc.)Technical b tA en em Execution re Ag Non Functional Scope of Agreement Functional Task/ Domain Gen. Common Data/ Purpose, Sense App * Industryabout Based it’s Broad the Info. business, not just computing resources
  9. 9. Outline• Semantics for Technical Services – Data Semantics * – Functional Semantics * – Non Functional Semantics * – Execution Semantics• Semantics for Knowledge Services• Conclusions*Can be represented using ontologies
  10. 10. Semantics for Technical Services Current and past focus of METEOR-S
  11. 11. Semantics for Technical Services • Data/Information Semantics – What: (Semi-)Formal definition of data in input and output messages of a web service – Why: for discovery and interoperability – How: by annotating input/output data of web services using ontologies • Functional Semantics – (Semi-) Formally representing capabilities of web service – for discovery and composition of Web Services – by annotating operations of Web Services as well as provide preconditions and effects • Execution Semantics – (Semi-) Formally representing the execution or flow of a services in a process or operations in a service – for analysis (verification), validation (simulation) and execution (exception handling) of the process models – using State Machines, Petri nets, activity diagrams etc. • Non Functional Semantics (WS-*) – (Semi-) formally represent qualitative and quantitative measures of Web process – Non- Quantitative includes security, transactions – Quantitative includes cost, time etc. – Business constraints and inter service dependencies (Domain and application ontologies)
  12. 12. Semantics for Technical Services Execution, Development Adaptation and / Description Mediation / Annotation WSDL, WSDL-S, BPWS4J, SAWSDL, WSMO, OWL-S activeBPEL, METEOR-S WSMX (MWSAF) METEOR-S BPEL, WS- (Semantic) UDDI Agreement, WS- Composition, METEOR-S Policy (MWSDI) METEOR-S Configuration Publication (MWSCF) and / Discovery Negotiation
  13. 13. Semantics for Technical Services Execution, Development Adaptation and / Description Mediation / Annotation WSDL, WSDL-S, BPWS4J, SAWSDL, WSMO, OWL-S activeBPEL. Data METEOR-S WSMX / Information (MWSAF) METEOR-S Semantics BPEL, WS- (Semantic) UDDIAgreement, WS- Composition, METEOR-S Policy (MWSDI) METEOR-S Configuration Publication (MWSCF) and / Discovery Negotiation
  14. 14. Semantics for Technical Services Execution, Development Adaptation and / Description Mediation / Annotation WSDL, WSDL-S, BPWS4J, SAWSDL, WSMO, OWL-S activeBPEL, METEOR-S WSMX (MWSAF) METEOR-S Functional Semantics BPEL, WS- (Semantic) UDDIAgreement, WS- Composition, METEOR-S Policy (MWSDI) METEOR-S Configuration Publication (MWSCF) and / Discovery Negotiation
  15. 15. Semantics for Technical Services Execution, Development Adaptation and / Description Mediation / Annotation WSDL, WSDL-S, BPWS4J, SAWSDL, WSMO, OWL-S activeBPEL, METEOR-S WSMX (MWSAF) METEOR-S Non Functional BPEL, WS- Semantics (Semantic) UDDIAgreement, WS- Composition, METEOR-S Policy (MWSDI) METEOR-S Configuration Publication (MWSCF) and / Discovery Negotiation
  16. 16. Semantics for Technical Services Execution, Development Adaptation and / Description Mediation / Annotation WSDL, WSDL-S, BPWS4J, SAWSDL, WSMO, OWL-S activeBPEL, Execution METEOR-S WSMX (MWSAF) Semantics METEOR-S BPEL, WS- (Semantic) UDDIAgreement, WS- Composition, METEOR-S Policy (MWSDI) METEOR-S Configuration Publication (MWSCF) and / Discovery Negotiation
  17. 17. Semantics for Technical Services Execution, Development Adaptation and / Description Mediation / Annotation WSDL, WSDL-S, BPWS4J, SAWSDL, WSMO, OWL-S activeBPEL, Data Execution METEOR-S WSMX / Information (MWSAF) Semantics METEOR-S Semantics Semantics Required for Web Processes QoS Functional Semantics Semantics BPEL, WS- (Semantic) UDDIAgreement, WS- Composition, METEOR-S Policy (MWSDI) METEOR-S Configuration Publication (MWSCF) and / Discovery Negotiation
  18. 18. DATA SEMANTICS
  19. 19. Data Semantics UDDI Query UDDI Registry Locate Suppliers Results Item Details Receive Quote Send Quote Request Quote Details Check Inventory Choose Supplier How does the supplier recognize Negotiate Agreement Item Details Negotiate Agreement Send Order Receive Order Supplier Process Customer Process
  20. 20. Data Semantics - options• Pre-defined agreement on all data fields – Limited flexibility, hard to integrate new suppliers in process• Use a standard like Rosetta Net/ebXML – Greater flexibility, but limited to suppliers following standard – Standard may not be expressive enough for everyones needs• Annotate data fields with domain ontologies – Most flexible, semi-automatic transformation based on ontology mapping – Ontology can be based on domain standard, while providing more flexibility and extensibility
  21. 21. WSDL-S Specification(Now the key input to W3C leading to Semantic Annotation of WSDL- SAWSDL)
  22. 22. PurchaseOrder.wsdls…………<xs:element name= " OrderConfirmation" type="xs:stringwssem:modelReference=" rosetta#PurchaseOrderResponse"/></xs:schema> Data from</types> Rosetta Net<interface name="PurchaseOrder"> Ontology<wssem:category name= “Electronics” taxonomyURI=http://www.naics.com/ taxonomyCode=”443112” /><operation name=“order” pattern=wsdl:in-out Function modelReference = "rosetta#RequestPurchaseOrder" > from Rosetta<input messageLabel = ”processPurchaseOrderRequest" Net Ontologyelement="tns:processPurchaseOrderRequest"/><output messageLabel ="processPurchaseOrderResponse"element="processPurchaseOrderResponse"/><!—Precondition and effect are added as extensible elements on an operation><wssem:precondition name="ExistingAcctPrecond"wssem:modelReference="POOntology#AccountExists"><wssem:effect name="ItemReservedEffect"wssem:modelReference="POOntology#ItemReserved"/></operation></interface>
  23. 23. Representing mappings <complexType name="POAddress"wssem:schemaMapping=”http://www.ibm.com/schemaMapping/POAdd Addressress.xsl#input-doc=doc(“POAddress.xml”)”><all> has_StreetAddress<element name="streetAddr1" type="string" /><element name="streetAdd2" type="string" /> xsd:string<element name="poBox" type="string" /><element name="city" type="string" /> has_City<element name="zipCode" type="string" /><element name="state" type="string" /> xsd:string<element name="country" type="string" /><element name="recipientInstName" type="string" /> has_Zip</all></complexType> xsd:stringWSDL complex type element OWL ontology Mapping using XSLT .... <xsl:template match="/"> <POOntology:Address rdf:ID="Address1"> <POOntology:has_StreetAddress rdf:datatype="xs:string"> <xsl:value-of select="concat(POAddress/streetAddr1,POAddress/streetAddr2)"/> </POOntology:has_StreetAddress > <POOntology:has_City rdf:datatype="xs:string"> <xsl:value-of select="POAddress/city"/> </POOntology:has_City> <POOntology:has_State rdf:datatype="xs:string"> <xsl:value-of select="POAddress/state"/> </POOntology:has_State>....
  24. 24. FUNCTIONAL SEMANTICS
  25. 25. Functional Semantics UDDI Query UDDI Registry Locate Suppliers Results Item Details Receive Quote Send Quote Request Quote Details Check Inventory Choose Supplier How to locate appropriate Negotiate Agreement supplier? Negotiate Agreement Send Order Receive Order Supplier Process Customer Process
  26. 26. Functional Semantics• Keyword based search in UDDI – Needs human involvement – Low precision and high recall• Port Type based search in UDDI – Requires service providers to agree on port types – Less flexible, requires total agreement on method names and data type names• Template Based Semantic Discovery – Requires ontological commitment of data types and operations – Can search on any or many aspects of description+interface – Can have complex similarity measures and be used to provide ranked results based on similarity
  27. 27. Semantic Templates Part of Rosetta Net Ontology• Semantic Templates capture the functionality of a Web service with the help of ontologies/other domain models• Find a service that sells RAM in Athens, GA. It must allow the user to return and cancel, if needed• The template can also have non- functional (QoS) requirements such as response time, security, etc. WSDL-S is used to capture semantic templates Data Semantics Functional Semantics Non-Functional Semantics
  28. 28. Semantic Discovery • Finds actual services matching semantic templates • Implemented as a layer over UDDI [1] • Current implementation based on ontological representation of operations, inputs and outputs • Returns ranked of services for each semantic template • Builds upon following previous discovery implementations – Extends matching presented in [2] to consider operations and service level metadata – Extends the approach presented “WSDL to UDDI Mapping” [3] to support operation level discovery[1] K. Verma, K. Sivashanmugam, A. Sheth, A. Patil, S. Oundhakar and John Miller, METEOR-S WSDI: A ScalableInfrastructure of Registries for Semantic Publication and Discovery of Web Services, JITM[2] M. Paolucci, T. Kawamura, T. Payne and K. Sycara, Semantic Matching of Web Services Capabilities, ISWC 2002.2[3] Using WSDL in a UDDI Registry, Version 2.0.2 - Technical Note, http://www.oasis-open.org/committees/uddi-spec/doc/tn/uddi-spec-tc-tn-wsdl-v202-20040631.pdf
  29. 29. Non Functional Semantics Business and Application constraints
  30. 30. Non Functional Semantics UDDI Query UDDI Registry Locate Suppliers Results Item Details Receive Quote Send Quote Request Quote Details Check Inventory Choose Supplier QoS Semantics Negotiate Agreement Negotiate Agreement Send Order Receive Order Supplier Process Customer Process
  31. 31. Non Functional Semantics• Does the supplier support customer’s business constraints – e.g. cost, supply time etc.• Interaction should adhere to the entities’ policies – e.g security, transactions• In case of more suppliers, domain constraints should be satisfied – e.g. a certain supplier’s parts do not work with other supplier’s parts
  32. 32. Non Functional Semantics• Used in lifecycle – Agreement Matching • Matching syntactically heterogeneous by semantically homogeneous agreements – Dynamic Process Configuration • Configuring process based on process constraint We will demonstrate how ontology-driven semantic approach supports these capabilities.
  33. 33. SWAPS: Use of Semantics in Agreement Matching An agreement is a collection of alternatives. A={Alt1, Alt2, …, AltN} An alternative is a collection of guarantees. Alt={G1, G2, ...GN} A guarantee is defined as a collection-“requirement(Alt, G) ” returns true if G is a requirement of G={Scope, Obligated, SLO, Qualifying Condition, BusinessAlt Value}“capability(Alt, G) ” returns true if G is an assurance of Alt“scope(G)” returns the scope of G“obligation(G) ” returns the obligated party and consumer There is a potential match between provider of G alternatives if:“satisfies(Gj, Gi)” returns true if the SLO of Gj is equivalenttoor stronger than the of one Gi For all requirement SLO of alternative, there is a capability in other alternative, which has the same scope and the sameAn alternative Alt1 SLO suitable match for Alt2 if:the request. obligation and the is a of the capability satisfies ("Gi) such that Gi ∈ Alt1 ∧ requirement(Alt1, Gi) ∧ ($Gj)
  34. 34. WS-Agreement Definition and Ontology hasGuaranteeTerm GuaranteeTerm An agreement consists of a collection of Guarantee hasBusinessValue terms hasScope A guarantee term has a scope – e.g. operation hasObjective hasCondition Scope BusinessValue of service Qualifying Condition ServiceLevelObjectivev hasReward Reward Predicate Predicate There might be business values hasPenalty associatedA guarantee term may have collection of qualifying A guarantee term may have a hasImportance condition for each guarantee terms. Business values withservice level objectives ParameterSLO’s to hold. Importance Parameter Penalty Value include importance, confidence, penalty, Unit ValueExpression e.g. numRequests Value and reward.e.g. responseTime < 2 seconds Unit < 100 ValueUnit e.g. Penalty 5 USD Assessment Interval ValueExpression OWL ontology Assessment Interval ValueUnit TimeInterval Count Count TimeInterval Agreement represented as an instance of ontology
  35. 35. SWAPS Ontologies  WS-Agreement: individual agreements are instances of the WS-Agreement ontology  Temporal Concepts: time.owl (OWL version of DAML time http://www.isi.edu/~pan/damltime/time.owl)  Concepts: seconds, dayOfWeek, ends  Quality of Service: Max Maximilien’s QoS ontology (IBM) -> Ont-Qos  Concepts: responseTime, failurePerDay  Domain Ontology: an ontology used to represent the domain
  36. 36. Using Semantic Agreements withWSDL-S WS-Agreement Ontology Agriculture Ontology Guarantee CropTime QoS Scope FarmerAdd BV r SLO Quality Pric Obligated e Predicate Split Moistur e Less WeightDomain Independent Greater Domain Dependent agri:moisture less 12% GetMoisture Adding Semantics to Agreements: Semantics to Web Services: Adding agri:splits less 20% obligated: less 12% GetSplits GetWeight Improves Monitoring and Negotiation agri:weight greater 54 lbs Enables more accurate discovery and composition. GetPrice Input: Address agri:priceWS-Agreement equals 10 USD Improves the accuracy of matching Merchant WS-Agreement Merchant Service WSDL-S
  37. 37. EvaluationConsumer Provider Approach 1: Approach 2: Approach 3: Approach 4:Requirement Capability Ontology Ontology Rules without No and Rules without Rules Ontologies Rules and No OntologyresponseTime responseTime < YES YES YES, but only if YES, but only if <5 4 parameters are parameters are named similar named similar syntactically syntacticallyresponseTime (duration1 + YES NO YES, but only if NO <5 duration2) the parameters <4 are named similar syntactically to the rule criteriaresponseTime rt < 4 YES YES NO NO <5responseTime networkTime < YES NO YES, but only if NO <5 2 the parameters executionTime are named similar <1 syntactically to the rule criteria
  38. 38. The Matching ProcessObligated: Provider Obligated: Provider99% of responseTimes < responseTime < 14 s14 s QC: day of week = weekday Consumer Penalty: 15 USD Provider1 Obligated: Provider Obligated: Provider failurePerWeek < 10 FailurePerWeek < 7 Penalty 10USD Obligated: Provider Obligated: Provider transmitTime < 4s failurePerWeek < 7 QC: maxNumUsers < 1000 Penalty: 2USD Penalty: 1 USD Provider2 Obligated: Provider ProcessTime < 5 s QC: numRequests < 500 Penalty: 1 USD
  39. 39. The Matching ProcessObligated: Provider Obligated: ProviderresponseTimes < 14 s responseTime < 14 s QC: day of week = weekday Consumer Penalty: 15 USD Provider1 Obligated: Provider Obligated: Provider failurePerWeek < 10 FailurePerWeek < 7 Penalty 10USD Knowledge from Domain Specific Rules: if (x >= 96) responseTime < y else responseTime > y
  40. 40. The Matching Process isEquivalentObligated: Provider Obligated: ProviderresponseTime < 14 s responseTime < 14 s QC: day of week = weekday Consumer Penalty: 15 USD Provider1 Obligated: Provider Obligated: Provider failurePerWeek <10 FailurePerWeek < 7 Penalty 10USDKnowledge from Semantics of Predicate Rules
  41. 41. The Matching ProcessObligated: Provider Obligated: ProviderresponseTime < 14 s responseTime < 14 s QC: day of week = weekday Consumer Penalty: 15 USD Provider1 Obligated: Provider Obligated: Provider failurePerWeek <10 FailurePerWeek < 7 Penalty 10USD isStrongerKnowledge from Semantics of Predicate Rules
  42. 42. The Matching ProcessObligated: ProviderresponseTime < 14 s Consumer Obligated: Provider failurePerWeek < 10 Obligated: Provider Obligated: Provider transmitTime < 4s failurePerWeek < 7 QC: maxNumUsers < 1000 Penalty: 2USD Penalty: 1 USD Provider2 Domain Specific Rule Obligated: Provider responseTime = transmitTime + processTime ProcessTime < 5 s QC: numRequests < 500 Penalty: 1 USD
  43. 43. The Matching ProcessObligated: ProviderresponseTime < 14 s Consumer Obligated: Provider failurePerWeek < 10 Obligated: Provider Obligated: Provider responseTime < 9s failurePerWeek < 7 QC: maxNumUsers < 1000 AND Penalty: 2USD numRequests < 500 Penalty: 1 USD Provider2
  44. 44. The Matching ProcessObligated: ProviderresponseTime < 14 s Consumer Obligated: Provider failurePerWeek < 10 isStronger Obligated: Provider Obligated: Provider responseTime < 9s failurePerWeek < 7 QC: maxNumUsers < 1000 AND Penalty: 2USD numRequests < 500 Penalty: 1 USD Provider2 isStrongerSteps #5-6: Comparison Rules
  45. 45. The Matching ProcessObligated: Provider notSuitable Obligated: ProviderresponseTime < 14 s responseTime < 14 s QC: day of week = weekday Consumer Penalty: 15 USD Provider1 Obligated: Provider Obligated: Provider failurePerWeek < 10 FailurePerWeek < 7 Penalty 10USD Obligated: Provider Obligated: Provider responseTime < 9s failurePerWeek < 7 QC: maxNumUsers < 1000 AND Penalty: 2USD numRequests < 500 Penalty: 1 USD User Preference Rule: Provider2 dayofWeek = weekday notSuitable
  46. 46. The Matching ProcessObligated: Provider Obligated: ProviderresponseTime < 14 s responseTime < 14 s QC: day of week = weekday Consumer Penalty: 15 USD Provider1 Obligated: Provider Obligated: Provider failurePerWeek < 10 FailurePerWeek < 7 Penalty 10USD Obligated: Provider Obligated: Provider responseTime < 9s failurePerWeek < 7 QC: maxNumUsers < 1000 AND Penalty: 2USD numRequests < 500 Penalty: 1 USD Provider2
  47. 47. Dynamic Process Configuration• Operations Research has been used in industry for business process optimization• There is often a lot of domain knowledge in business process optimization – Minds of analysts/experts – Hidden in databases/texts• We try to explicitly capture domain knowledge and link with IT systems
  48. 48. Dynamic Process Configuration Find optimal partners for the process based on process constraints – cost, supply time, etc. Conceptual Approach 1. Create framework to capture represent domain knowledge 2. Represent constraints on the domain knowledge 3. Ability to reason on the constraints and configure the process
  49. 49. Dynamic Process Configuration Research Challenges – Capturing functional and non-functional requirements of the Web process (Abstract process specification) – Discovering service partners based on functional requirements (Semantic Web service discovery) – Choosing optimal partners that satisfy non- functional requirements (Constraint Analysis)K. Verma, R. Akkiraju, R. Goodwin, P. Doshi, J. Lee, On Accommodating Inter Service Dependencies in Web Process Flow,AAAI Spring Symposium on Semantic Web Services, 2004R. Aggarwal, K. Verma, J. A. Miller, Constraint Driven Composition in METEOR-S, SCC 2004.K. Verma, K.Gomadam, J. Miller and A. Sheth, Configuration and Execution of Dynamic Web Processes, LSDIS Lab Technical Report, 2005.
  50. 50. Abstract Process Specification 1. Specify process control flow by using virtual partners 2. Specify Process Constraints 3. Capture Functional Requirements of Services using Semantic Templates
  51. 51. Process Constraints• Constraints can be specified on a partner, an activity or the process as a whole.• An objective function can also be specified e.g., minimize cost and supply-time, etc.• Two types of constraints: – Quantitative (Q) (Time < 5 sec) – Logical (L) (preferredPartner, Security, etc.)
  52. 52. Process Constraints Feature Scope Goal Value Unit AggregationCost (Quantitative) Process Minimize Dollars ΣSupplytime (Quantitative) Process Satisfy <7 Days MAXCost (Quantitative) Activity Satisfy <200000 Dollars ΣPreferredSupplier(P1) Partner 1 Satisfy True (Logical)Compatible (P1, P2) Process Satisfy True (Logical)
  53. 53. Constraint Analysis• Multi-paradigm proposed: – Integer Linear Programming for quantitative constraints – Semantic Web Rule Language and OWL for domain constraints• Discovered Services first given to ILP solver – It returns ranked sets of services• Then each set is checked for logical constraints using a SWRL reasoner – Sets not satisfying the criteria are rejected
  54. 54. Domain Ontology – Detailed View
  55. 55. Rules• Supplier 1 should be a preferred supplier. – “if S1 is a supplier and its supplier status is preferred then the S1 is a preferred supplier”. Supplier (?S1) and partnerStatus (?S1, “preferred”) => preferredSupplier (?S1)• Supplier 1 and supplier 2 should be compatible. – if S1 and S2 are suppliers and they supply parts P1 and P2, respectively, and the parts work with each other, then suppliers S1 and S2 are compatible for parts P1 and P2. Supplier (?S1) and supplies (?S1, ?P1) and Supplier (?S2) and supplies (? S2, ?P2) and worksWith (?P1, ?P2) => compatible (?S1, ?S2, ?P1, ?P2) RAM (?P1) and MB (?P2) and worksWithMB (?P1, ?P2) =>worksWith (? P1, ?P2)
  56. 56. Configuration Step 1: SemanticDiscovery
  57. 57. Configuration Step 2: QuantitativeConstraint Analysis
  58. 58. Configuration Step 3: Logical ConstraintAnalysis
  59. 59. EXECUTION SEMANTICS
  60. 60. Execution Semantics UDDI Query UDDI Registry Locate Suppliers Results Item Details Receive Quote Send Quote Request Quote Details Check Inventory Choose Supplier Execution Semantics 1. How to recover from Negotiate Agreement Negotiate Agreement physical/ logical errors (e.g. delays in goods) Send Order Receive Order Supplier Process Customer Process
  61. 61. Process Adaptation• Ability to adapt the processes from failures, unexpected events• Two kinds of failures – Failures of physical components like services, processes, network • Can replace services using dynamic configuration – Logical failures like violation of SLA constraints/Agreements such as Delay in delivery, partial fulfillment of order • Need additional decision making capabilitiesK. Verma, A. Sheth, Autonomic Web Processes, ICSOC 2005K. Verma, P. Doshi, K. Gomadam, A. Sheth, J. Miller, Optimal Adaptation of Web Processes with Coordination Constraints, ICWS 2006.
  62. 62. Process AdaptationAdaptation Problem Optimally react to events like delays in ordered goods Conceptual Approach 1. Maintain states of the process – normal states, error states, goal states 2. Capture costs while transitioning from error states to goal state 3. Ability to decide optimal actions on the basis of stateK. Verma, A. Sheth, Autonomic Web Processes, ICSOC 2005K. Verma, P. Doshi, K. Gomadam, A. Sheth, J. Miller, Optimal Adaptation of Web Processes with Coordination Constraints, ICWS 2006.
  63. 63. Marginalizing events
  64. 64. Generating States using preconditionsand effects Actions Chance Variables Events
  65. 65. Generated State Transition DiagramState Values of Boolean Explanation No. variables1 Ordered <O C R Del Rec >2 Ordered and Canceled <O C R Del Rec >3 Ordered and Delayed <O C R Del Rec >4 Ordered, Received and <O C R Del Rec > Returned5 Ordered, Delayed and <O C R Del Rec > Cancelled6 Ordered, Delayed, Received <O C R Del Rec > and Returned7 Ordered, Delayed and <O C R Del Rec > Received8 <O C R Del Rec > Ordered and Received
  66. 66. Costs and Probabilities• Costs of ordering taken from configuration module – From first two service sets • Optimal supplier and alternate supplier• Probability of delay and cost of returning and canceling taken from supplier policy – Can be represented using WS-Policy or WS- Agreement
  67. 67. Supplier Policy – The supplier gives a probability of 55% for delivering the goods on time. – The manufacturer can cancel or return goods at any time based on the terms given below. • If the order is delayed because of the supplier, the order can be cancelled with a 5% penalty to the manufacturer. • If the order has not been delayed, but it has not been delivered yet, it can be cancelled with a penalty of 15% to the manufacturer. • If the order has been received after a delay, it can be returned with a penalty of 10% to the manufacturer. • If the order has been received without a delay, it can be returned with a penalty of 20% to the manufacturer.
  68. 68. Costs and Probabilities Current State Action Next State Cost<O C R Del Rec > NOP <O C R Del Rec > 0<O C R Del Rec > CANCEL <O C R Del Rec > 150<O C R Del Rec > DEL <O C R Del Rec > 0<O C R Del Rec > RECEIVE <O C R Del Rec > 0<O C R Del Rec > ORDER <O C R Del Rec > 100<O C R Del Rec > NOP <O C R Del Rec > DelayCost = {200, 300, 400}<O C R Del Rec > CANCEL <O C R Del Rec > 50<O C R Del Rec > RECEIVE <O C R Del Rec > 0<O C R Del Rec > ORDER <O C R Del Rec > 100<O C R Del Rec > ORDER <O C R Del Rec > 100<O C R Del Rec > ORDER <O C R Del Rec > 100<O C R Del Rec > CANCEL <O C R Del Rec > 150<O C R Del Rec > NOP <O C R Del Rec > 0<O C R Del Rec > RETURN <O C R Del Rec > 200<O C R Del Rec > NOP <O C R Del Rec > 0
  69. 69. Handling Coordination Constraints• Since the RAM and Motherboard must be compatible, the actions of service managers (SMs) must be coordinated• For example, if MB delivery is delayed, and MB SM wants to cancel order and change supplier, the RAM SM must do the same• Hence, coordination must be introduced in SM- MDPs
  70. 70. Centralized Approach• State space created by Cartesian product of transition diagrams• Joint actions from each state• Transition probability created by multiplying states• Costs created by adding cost per action from each state – Compatible actions given rewards – Incompatible actions given penalties• Optimal but exponential with number of manager
  71. 71. Decentralized Approach• Simple coordination mechanism• If one service manager changes suppliers – All dependent managers must change suppliers• Low complexity but sub- optimal
  72. 72. Hybrid Approach• If the policy of some SM dictates it to change suppliers, the following actions happen: – it sends a coordinate request to PM – PM gets the current cost of changing suppliers or current optimal action by polling all SMs• It takes the cheapest action (change supplier or continue)• A bit like decentralized voting- will change suppliers if majority are delayed• It mirrors performance of centralized approach and has complexity like the decentralized approach
  73. 73. Evaluating Process Adaptation• Evaluation with the help of the supply chain scenario• Two main parameters used for the evaluation – Probability of Delay – (probability that an item ordered from a supplier will be delayed) – Penalty of Delay – (cost for the manufacturer for not reacting to delay)• Total process cost = $1000 and cost of changing suppliers (CS) =$200
  74. 74. Evaluating Adaptation
  75. 75. Semantics for Lightweight Services
  76. 76. Lightweight services and Mashups• REST based implementation becoming popular – SOAP -> Web service – REST -> Lightweight Web service• REST services exposed as API’s – Eg. Google Maps API, Flickr API• Mashups combine information from different services on the Web to create services with additional value• Asynchronous Javascript And XML (AJAX) is primarily used by mashups to display the results to the user
  77. 77. Current limitations and Role ofsemantics• Current Mashups tightly coupled (lack dynamism) – E.g. HousingMaps.com uses craigslist and Google maps.• Tight binding limits effectiveness – Better information may be available for a specific area – E.g. for Atlanta area, realtor1.com might be a better service than craigslist.• Can annotate XML for automated integration
  78. 78. An example• Consider a mashup: mybook.com – Allows users to search and buy used and new books – Gets data from various vendors on the web• Can customize vendors based on requests – E.g., discover two vendors, ubn.com and yaos.com on the fly• Use conceptual model/ontology based annotation of XML data for integration – mybook.com can interpret the XML documents from vendors with help of annotations
  79. 79. An Example of Smashup (Semantic mashup)
  80. 80. Semantics for Knowledge Services Current and past focus of METEOR-S
  81. 81. Semantics for Knowledge Services• Work in last two decades on knowledge modeling not so successful – Focus on capturing knowledge – However most businesses use people to solve problems not expert systems• Knowledge service try to create semantic profiles of human expertise – Focus on “who can” not “how to” – Use of ontologies for shared descriptions
  82. 82. High Level Model for KnowledgeServices
  83. 83. Using Model for Knowledge Services• Such a model can be used to answer questions – Find managers who have led project worth at least a million dollars – Find developers who have created multimedia services using Java – Find consultants who have some expertise in Law
  84. 84. Autonomic Web Processes• The goal (Albatross) – Self Configuring, Self Healing, Self Optimizing, Self Protecting Business Processes• Realization – Comprehensive modeling of business processes using 3S model• Advantages – Alignment of technology with business goals – Dynamic processes that adapt with the changing environment
  85. 85. Conclusions• Businesses perceive IT as an extension of business strategy – 3S Model uses semantics to provide a comprehensive model of human and technical assets – Modeling and exploitation of four types of semantics• CS Researchers must take a more pervasive view of services

×