cool: a context ontology language to enable contextual interoperability thomas strang, claudia...

Download CoOL: A Context Ontology Language to Enable Contextual Interoperability Thomas Strang, Claudia Linnhoff-Popien, and Korbinian Frank German Aerospace Centor

Post on 17-Jan-2016

213 views

Category:

Documents

0 download

Embed Size (px)

TRANSCRIPT

  • CoOL: A Context Ontology Language to Enable Contextual InteroperabilityThomas Strang, Claudia Linnhoff-Popien, and Korbinian FrankGerman Aerospace Centor (DLR). Oberpfaffenhofen, GermanyLudwig-Maximilians-University (LMU), Munich, Germany

    Presented by Sangkeun Lee IDS Lab.

    Center for E-Business Technology

    Copyright 2008 by CEBT

    IntroductionThe pervasive computing trend is driving a need for context-aware service architectures involved in a service interactionTo reduce the amount of required user actionsVarious actorsE.g. any user, any service provider, environment, or third partiesA key to context information in context-aware system is a Well designed model to describe contextual facts and contextual interrelationshipsPrevious works lack formalityThe context modeling approach in this paper tries to close the formality gap by using OntologiesAnother example: CONON (Context Ontology)IDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    What is Context?Definition of Context Information A context information is any information which can be used to characterize the state of an entity concerning a specific aspect An entity is a person a place or in general an object An aspect is a classification, symbol or value-range, whose subsets are a superset of all reachable states IDS Lab. Seminar - *Center for E-Business TechnologyReference: Trends in Mobile Computing From Mobile Phone to Context-Aware Service Platform by Thomas Strang

    Center for E-Business Technology

    Copyright 2008 by CEBT

    What is Context? (contd)Definition of Context A context is the set of all context information characterizing the entities relevant for a specific task in their relevant aspects An entity is relevant for a specific task, if its state is characterized at least concerning one relevant aspect An aspect is relevant, if the state with respect to this aspect is accessed during a specific task or the state has any kind of influence on the task IDS Lab. Seminar - *Center for E-Business TechnologyReference: Trends in Mobile Computing From Mobile Phone to Context-Aware Service Platform by Thomas Strang

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Context-awarenessDefinition of Context Awareness: A system is context aware if it uses any kind of context information before or during service provisioning or service usage Two main benefits from Context Awareness: Adaptation of services to changes in environment reduces amount of interaction with user Improvement of UI (particularly on small mobile devices)IDS Lab. Seminar - *Center for E-Business TechnologyReference: Trends in Mobile Computing From Mobile Phone to Context-Aware Service Platform by Thomas Strang

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Ontologies and the Context Ontology LanguageChallenge to describe contextual facts and interrelationships in a precise and traceable mannerprint document on printer near to meIt is required to have a precise definition of terms used in the task, particularly what near means to meIt is highly desirable that each participating party in a service interaction share the same interpretation of the meaning behind itShared understandingOntologiesMay be stored at different places and created by different authors, which offers flexibility and extensibilityIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Context Ontology Language (CoOL)CoOL CoreOWL and DAML+OILF-LogicCoOL IntegrationA collection of schema and protocol extensions as well as common subcontepts, enabling the use of CoOL Core in several service frameworks (e.g Web Service)Out of focus of this paperHaving a projection of the model in multiple ontology languages enables thatDeveloper may use any of languages which seems to be adequateUsing OWL because of wide range of available toolsUsing F-logic because of its rule based extensibility

    IDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    ASC ModelAspect-Scale-Context (ASC) modelNamed after the core conceptsEach aspect aggregate one or more scalesEach scale aggregates one or more context informationhasAspect, hasScale, constructedByIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    ASC Model ExamplesIDS Lab. Seminar - *Center for E-Business TechnologyReference: Ubiquitous Computing :Context and Context-Awareness, University of Innsbruck, Lecture slide from SS 2005 Another Example:SpartialDistanceAspect - MeterScale,KilometerScale 10, 20Fahrenheit's temperature scaleCelsius temperature scaleSpecific Values

    Center for E-Business Technology

    Copyright 2008 by CEBT

    OperationsIntraOperationMapping function from one scale to at least one other of the already existing scales of the same aspectInterOperationScales which require access to scales of one or more other aspects can be defined using InterOperationsE.g.KilometerPerHourScale of a SpeedAspect can be defined by using an InterOperation with two Parameter, delta_s and delta_t wheredelta_s is from an aspect SpatialDistanceAspectdelta_t is from an aspect DurationAspectMetricOperationCompare two context information instancesIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Transfer ModelASC model may be used as transfer model to employ the knowledge expressed in other context modelsExample: HenricksenContext extension to the Object-Role Modeling approachThe basic modeling concept in ORM is factHenricksen extended ORM to allow fact types to be categorized either as static or dynamic (profiled, sensed, derived)Using ASC ModelFacts can be modeled as context informationClassification can be mapped by introducing a quality aspect consisting the element {static, dynamic profiled, dynamic sensed, dynamic derived}dependOn relation can be expressed using Intra/InterOperationsIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Relation to DAML-SDAML-S : ontology of servicesServiceProfile, ServiceModel, and ServiceGroundingExtension: ServiceContextMay serve as a formal description of a services contextual interoperability by providing a comprehensive but extensible model based on the ASC modelIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Motivation: MNM Service Model and the Context ExtensionIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    Copyright 2008 by CEBT

    System ArchitectureIDS Lab. Seminar - *Center for E-Business TechnologyThe authors focus on the context provider domain as introduced in the previous page

    Center for E-Business Technology

    Copyright 2008 by CEBT

    Discussion & ConclusionsDiscussionASC modelThe authors gives us a different point of view (compare to CONON)Whats weak points and strong points?It is interesting the authors classified actors intoContext Providers, Service Providers, CustomersWho can be Context Providers?ConclusionsThe authors introduced ASC model as a base model to express how some context information can be used to characterize the state of an entity concerning a specific aspectASC model fits into a general purpose service modelContext Extension : Making any service interaction based on that model context-awareASC model as transfer modelIDS Lab. Seminar - *Center for E-Business Technology

    Center for E-Business Technology

    *