320 likes | 488 Views
Resource Description Framework and Ontology. Cungen Cao Knowledge Acquisition & Sharing Group Institute of Computing Technology Chinese Academy of Sciences March 25, 2002. AI Ontology. T. Gruber’s Definition. An ontology is a formal explicit specification of a shared conceptualization.
E N D
Resource Description Framework and Ontology Cungen Cao Knowledge Acquisition & Sharing Group Institute of Computing Technology Chinese Academy of Sciences March 25, 2002
T. Gruber’s Definition An ontology is a formal explicit specification of a shared conceptualization
Necessary Conditions for “Something” to Be Called an Ontology • Exhaustive • Everything must be stated • Explicit • Everything must be stated explicitly • Formal • Everything must be stated formally • Consensual • Everything stated is consensus
defcategory Basic-Entity { attribute: full-name :type String attribute: abbreviated-name :type String attribute: informal-definition :type Text attribute: formal-definition :type First-Order-Formula relation: is-instance :type Concept relation: has-part :type Instance …… } Geographical Ontology
defcategory Location { relation: in-east-of :type PhysicalEntity relation: in-south-of : type PhysicalEntity relation: in-west-of : type PhysicalEntity relation: in-north-of : type PhysicalEntity relation: in-middle-of : type PhysicalEntity }
defcategory Region inherits Basic-Entity; includes Location { attribute: population :type Integer :facet time attribute: land-area :type Real :unit km2 attribute: average-summer-whether :type Real :unit ℃ … }
defcategory City inherits Region { relation: geographical-part-of :type GeographicalEntity attribute: city-flower :type PlantEntity … } defcategory Country inherits Region { attribute: capital-city :type GeographicalEntity attribute: first-regional-division :type GeographicalEntity attribute: national-flower :type GeographicalEntity attribute: number-of-nationalities :type Integer … }
Axiom 1. geographical entities X and Y, in-east-of(X,Y) in-west-of(Y,X). Axiom 1. geographical entities X and Y, in-east-of(X,Y) in-west-of(Y,X) in-northwest-of (Y,X) in-southwest-of (Y,X). Axiom 2. city X, country Y, geographical-part-of(X,Y) population(X) population(Y). Axiom 3. city X country Y, geographical-part-of(X,Y) number-of-nationalities(X) number-of-nationalities(Y). Axiom 4. city X, country Y, capital-city(X,Y)geographical-part-of(X,Y) Geographical Axioms
Steps in Ontology Building • Which domain? • What purposes? • What categories? • What attributes for each category? • How are categories interrelated? • How are relationships constrained?
Standard Universal Ontology? • Is there any universal ontology covering everything? • Most probably no! • Every enterprise has his own preferences • Ontology translation is a necessity
Web Information Retrieval • Agent-based Search Engines • E.g. Altavista, Infoseek, Excite • Brute force • Metadata-based Search Engines • E.g. Yahoo • Man-generated categories & subject labels
Metadata-Based Information Retrieval • Metadata • Book • Author, Title, Publisher, Date of Publication, Subject • lookup information: Bai Shuo, Li Guojie • Person • Name, Surname, First Name, Gender, Date of Birth, Date of Place, etc. • Geographic Entity • Name, Location, Population, Nationalities, etc.
Resource Description Framework • RDF is a common framework for representing metadata for Web resources between different applications • RDF descriptions can be exchanged without loss of semantics! • A resource is anything that can have a URI • E.g. my electronic paper Ontology whose URL is http://www.nki.net.cn/ontology.htm
Resource Description Framework • Resources are specified by properties • Title, author, date of last modification, etc. • See Dublin Core Metadata • A property is a resource having a name, URI, and it own properties • You can define your own properties! • Properties defined by you can be searched, manipulated and sold like any other resource!
Resource Description Framework • A statement (Resource, Property, Value) or (Subject, Predicate, Object) is a resource • You can define metadata for statements!
Why Is XML Better Than RDF? • Advantages of XML • A device for inventing tags freely to structure your documents • E.g. The company <name> IBM </name> …
Why Is XML Better Than RDF? • Disadvantages of XML • XML is a serial language, Maintaining such an order over Internet may not be time-consuming • When you represent general XML documents in computer memory, you get weird data structures that mix trees, graphs, and character strings. In general, these are hard to handle in even moderate amounts, let alone by the billion
XML Is the Foundation of RDF • RDF relies on the support of XML, and XML syntax is only one possible syntax for RDF • RDF uses XML to exchange descriptions of resources RDFS RDF XHTML XML HTML
Is RDF Enough? • RDF itself does not have an inference engine • the RDF data model does not provide any mechanism for declaring specific classes of resources, or for declaring which properties may be applicable to which classes of resources.
We Need RDF Schemas • A Type System • An RDF schema defines the properties of a resource (title, author, subject, size, color, etc.) • An RDF schema defines the types of resources (e.g. Web pages, animals and geographical entities) which the properties can be applicable. • An RDF schema defines properties in terms of the classes of resource to which they apply. • Domain and range constraints on the properties
RDF Schemas • RDFS Classes • rdfs: Class • rdfs: Resource • rdf: Property • rdfs:ContrainstProperty • rdfs: Literal
RDF Schemas • RDF Properties (Instances of rdf: Property) • rdf: type. (R, rdf:type, C)R is an instance of C • rdfs: subClassOf. (C, rdfs: subClassOf, rdf:type rdfs: Class) • rdfs: subPropertyOf • rdfs: domain • Domain constraints define the applicability of properties • rdfs: range • Range constraints are only applied to properties • Value of a range constraints is a class
rdfs: label • rdfs: comment • rdfs: seeAlso • rdfs: isDefinedBy • rdfs: subPropertyOf rdfs: seeAlso
OIL: Ontology Inference Layer • An ontology infrastructure for semantic Web • Vrije, Manchester, Stanford, Bell lab, … • Frame Formalism + Description Logic • For OIL ontology semantics • Ontology design reasoning support • (FaCT by Manchester)
OIL Layered Architecture • Core OIL • Compatible with RDFS • Standard OIL • Heavy OIL
Gene Ontology • GO (Gene Ontology) is a controlled vocabulary used to describe the biology of a gene product in any organism. • A gene product has one or many molecular functions • A gene product may be used in one or many biological processes • A gene product may appear in several cellular component
Gene Ontology • Purpose: Common Language for Annotating Gene Products • Useful Internet Resources • http://www.geneontology.com/ • http://genome-www.stanford.edu/GO • http://www.ebi.ac.uk/interpro/QuickGo
Gene Ontology • GO has 3 independent ontologies in DAG • molecular function describes tasks done by gene products at molecular level • biological process describes broad biological goals accomplished by ordered assemblies of molecular functions • cellular component describes subcellular structures, locations, and macromolecular complexes