C-OWL

Document Sample
C-OWL Powered By Docstoc
					Enhancing ontology integration
with contextualization

                     Andrei Tamilin



                      University of Trento
                             July 2, 2003
Outline                                                                                 2

 Introduction:
  Ontology & The Semantic Web in a glance
 Web Ontology languages
 Logical foundations of OWL semantics
 Contextualizing ontology semantics
 C-OWL language
 Conclusions
 References
          Assignment for Logics for knowledge representation and reasoning ICT course
                                                                                   3



          Introduction:
Ontology & The Semantic Web in a
              glance



     Assignment for Logics for knowledge representation and reasoning ICT course
…Ontology                                                                                 4

 From Aristoteles times                                               384 BC – 322 BC

   “Ontologia” is that part of Metaphysics that specifies the
    most fundamental categories of existence, the elementary
    substances or structures out of which the world is made


 Recently term “Ontology” has been taken up by
  researchers in Artificial Intelligence
   Building blocks out of which models of the real world are
    made
   Basic level of knowledge representation scheme


            Assignment for Logics for knowledge representation and reasoning ICT course
What is an Ontology?                                                                       5

 Ontology is a formal, explicit specification of a
  shared conceptualization (Tom Gruber)

   “Formal” refers to the fact that the ontology should be
    machine understandable, i.e. the machine should be able
    to interpret the semantics of the information provided
   “Explicit” means that the type of concepts used, and the
    constraints on their use are explicitly defined
   “Shared” means that an ontology captures consensual
    knowledge, that is, it is not restricted to some individual, but
    accepted by a group
   “Conceptualization” refers to an abstract model of
    phenomena in the world by having identified the relevant
    concepts of those phenomena

             Assignment for Logics for knowledge representation and reasoning ICT course
The Semantic Web vision                                                                6

The Semantic Web is an extension of the
current web in which the information is given
well-defined meaning, better enabling
computers and people to work in cooperation

  Tim Berners-Lee, James Hendler and Ora Lassila
  The Semantic Web, Scientific American, May 2001




         Assignment for Logics for knowledge representation and reasoning ICT course
Adding meaning to the resources                                                       7


                                     What computers understand?

                                     “Blah blah blah <a href=…> blah blah blah”

                                     Way of describing information meaning
                                     in the machine-processable format
                                     is required!



      Current Web                                        The Semantic Web




        Assignment for Logics for knowledge representation and reasoning ICT course
Ways of describing semantics                                                            8

 Two major broad approaches are exist:
   Ontologies – shared model of some domain that
    encode a view common to set of different parties
   Contexts – local models encoding a one party’s
    view of the domain
   Both approaches has strengths and weaknesses


 Current vision of the Semantic Web is
  following the first approach

          Assignment for Logics for knowledge representation and reasoning ICT course
Emerging tasks dealing with ontologies 9

 Having a one global ontology for the describing of
  everything possibly could be a good idea, but this is
  ideal situation not realized in the dynamic and
  chaotic Web

 2 meaning of the ontology integration task:
   Integration of ontologies is needed when building a new
    ontology reusing other available ontologies
   Integration of ontologies by merging different ontologies
    about the same subject into a single one that “unifies” them


            Assignment for Logics for knowledge representation and reasoning ICT course
                                                                               10




Web Ontology languages




 Assignment for Logics for knowledge representation and reasoning ICT course
Web language standards                                                                       11
 eXtensible Markup Language – machine-readable format of data
  representation
 The Resource Description Framework (RDF) – gives a mean for
  adding semantics
 RDF Schema (RDFS) – basic modeling language. Gives the means
  for describing
     concepts
     properties
     Is-A hierarchy
     Simple domain and range restrictions

 Requirements for web ontology language:
     Compatible with existing Web standards (XML, RDF, RDFS)
     Easy to understand and use
     Formally specified and of “adequate” expressive power
     Possible to provide automated reasoning support



               Assignment for Logics for knowledge representation and reasoning ICT course
OIL, DAML-ONT, DAML+OIL and OWL                                                              12
 Two languages developed to satisfy above requirements
   OIL: developed by group of European researchers
   DAML-ONT: developed in DARPA DAML program
 Efforts merged to produce DAML+OIL
    Submitted to W3C as basis for standardisation

 WebOnt working group developing OWL language standard
 DAML+OIL/OWL “layered” on top of RDFS
   RDFS based syntax and ontological primitives (subclass etc.)
   Adds much richer set of primitives (transitivity, cardinality, . . . )
   Describes structure of domain in terms of Classes and Properties




               Assignment for Logics for knowledge representation and reasoning ICT course
                                                                                    13



Logical foundations of OWL semantics




      Assignment for Logics for knowledge representation and reasoning ICT course
OWL basics                                                                                 14

 OWL family:
   OWL Lite – easy implementation with the functional subset
    allowed
   OWL DL – implements DL constructs satisfying computational
    properties of reasoning systems
   OWL Full – relaxes some of the constraints of OWL DL language
    therefore violating constraints of existing DL reasoners

 OWL DL based on the SHIQ description logic and therefore
  benefits:
   Well defined semantics
   Known reasoning algorithms
   Highly optimized implemented reasoning systems


             Assignment for Logics for knowledge representation and reasoning ICT course
Constructors of OWL language                                                         15




       Assignment for Logics for knowledge representation and reasoning ICT course
Global semantics of OWL                                                                    16

 OWL ontology is a set of axioms and facts, plus
  import references to other ontologies

 OWL space is a set of all ontologies

 Characteristic feature of OWL – global semantic
  approach
   There is a global interpretation
   Interpretation satisfies the OWL space if it satisfies each
    axiom and fact of each ontology
   Language is interpreted against a global domain

             Assignment for Logics for knowledge representation and reasoning ICT course
Desirable but not achievable                                                                 17
capabilities of OWL
 During the ontology integration process we would like to support:

   Directionality of information flow
       The flow from the target ontology to the source one is undesirable

   Local domains
       Give up the hypothesis that all ontologies are interpreted in a single
       global domain

   Context mappings
       The possibility of relating extensionally different elements of two
       ontologies, but contextually similar



               Assignment for Logics for knowledge representation and reasoning ICT course
                                                                                   18



Contextualizing ontology semantics




     Assignment for Logics for knowledge representation and reasoning ICT course
Dealing with directionality                                                                  19

 Idea: split global interpretation to a family of local interpretations,
  one for each ontology
 But: we still consider the global domain

 Introduce a new interpretation called “Hole”, that verifies any set
  of axioms and facts, possibly contradictory

 Interpretation satisfies OWL space if for each axiom and fact of
  each ontology there exist local interpretation


 Important observation:
  Keeps OWL syntax unchanged, modifies the OWL semantics

               Assignment for Logics for knowledge representation and reasoning ICT course
Dealing with local domains                                                               20

 Idea: associate to each ontology a local domain
 Each ontology interpreted by local interpretation on
  its local domain

 Restriction: any foreign expression on the ontology
  should be interpretable in its local domain


 Important observation:
  Keeps OWL syntax unchanged, modifies the OWL
  semantics


           Assignment for Logics for knowledge representation and reasoning ICT course
Dealing with context mappings                                                             21

 Context mapping allows us to state that a certain
  property holds between two different but
  contextually equivalent elements of two ontologies
 Mapping between two ontologies is a set of bridge
  rules between them (mappings are directional)
   With mappings ontology translates another ontology into its
    own (local) semantics

 Important observation:
  Requires an extension of OWL syntax (constructs
  for representing bridge rules must be added)

            Assignment for Logics for knowledge representation and reasoning ICT course
                                                                              22




       C-OWL language




Assignment for Logics for knowledge representation and reasoning ICT course
C-OWL intuition                                                                        23

 C-OWL syntactically is OWL with a set of
  mappings expressing the bridge rules,
  replicated from CtxML

 Benefits of C-OWL:
  Full representation power of OWL
  Full representational power of CtxML



         Assignment for Logics for knowledge representation and reasoning ICT course
C-OWL example                                                                             24
                   Wine                                   Vino
WhiteWine


                                                                   VinoNero




                                                                 VinoRosato


                                                            VinoBianco

  RedWine
                     Teroldego




            Assignment for Logics for knowledge representation and reasoning ICT course
                                                                              25




                Conclusions




Assignment for Logics for knowledge representation and reasoning ICT course
                                                                                        26

 Semantics of ontologies is fixed in order to
  satisfy desired requirements of ontology
  integration process

 Intuition of new extensional syntax to deal
  with new semantics is proposed




          Assignment for Logics for knowledge representation and reasoning ICT course
                                                                                       27
 Existing syntax and semantics of OWL was
  extended to deal with some problems that
  couldn’t otherwise be dealt with
 Extended language Context OWL is
  proposed to allow ontologies localize their
  contents.




         Assignment for Logics for knowledge representation and reasoning ICT course
References                                                                              28

 P.Bouquet, F.Giunchiglia, F. van Harmelen,
  L.Serafini, H.Stuckenschmidt. C-OWL:
  Contextualizing ontologies.
 A.Borgida, L.Serafini, Distributed description
  logics: preliminary investigations
 I.Horrocks, Logical foundations for the
  Semantic Web.



          Assignment for Logics for knowledge representation and reasoning ICT course

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:2/15/2012
language:English
pages:28