Presentation on theme: "ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 1 Disseminating Web (2.0) information XML standards : RDF, OWL Henry Boccon-Gibod (EDF R&D) Tuesday, November."— Presentation transcript:
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 1 Disseminating Web (2.0) information XML standards : RDF, OWL Henry Boccon-Gibod (EDF R&D) Tuesday, November 20, morning : 8:00 – 12:00
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 2 Disseminating (Web 3.0 ? ) knowledge XML standards : RDF, RDFS, OWL Henry Boccon-Gibod (EDF R&D) Tuesday, November 20, morning : 8:00 – 12:00
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 3 Something anybody may observe
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 4 Something anybody may observe Human people speak together using two main kinds of expressions : –Messages, Orders, Documents, Structured, sequentially organized, in a way already known by both –speaker and auditor –writer and reader –Descriptions set of assertions with relationships not significantly ordered, but organized –A grammar is a description –Schematics are descriptions Are the way knowledge is exchanged Obviously that is not obvious : –Messages, orders, documents may encapsulate descriptions –Descriptions may refer message, orders and documents Messages, Orders, Documents need initial Descriptions… …Of the way they are structured, sequentially organized
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 5 Trees and Graphs Typically –Messages, orders, documents are organized as trees Standards for documents such as GML, SGML and then XML provide basic structures to express tree organizations –Descriptions are organized as graphs Linguists, terminologists, knowledge engineers –are claiming for graph expressions –cannot be fully satisfied by XML structures designed for documents. Obviously –Document Type Definitions, XML schemas may express descriptions Unbounded Choices, ID, IDrefs mechanisms allows graphs expressions XML Schema, XSL specifications are kind of graph descriptions… But are not suitable to actually support efficient knowledge expression Hopefully –Any graph may be supported by a tree…
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 6 From Syntaxic Web To Semantic Web
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 7 Djoser Pyramid, designed by Imhotep
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 8 XML Schémas HTML DTDs Web Pyramid, designed by Sir Tim Berners Lee
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 9 URIUnicodeHTML … the earth of computers was an hostile wasteland, and darkness covered the abyss of operating systems, while a mighty wind swept over the proprietary networks” “ then Tim Berners Lee said, "Let there be the Web," and Robert Cailleau did it, and there was the Web… ” “… and the Web (1.0) was just a way to share HTML published identified documents stored on heterogeneously designed computers linked together by network of networks using the Internet protocol, and to display them through navigator software “ “In the beginning…
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 10 URIUnicode NamespacesXML DTDsXML Schémas HTML Then came new recommendations… Of structuring principles adapted for all kinds of documents –Defined using structures of embedded segments, delimited by tags, –XML, adaptation of SGML, which HTML is one model (DTD)
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 11 LANGUE ART TI SEC1 TSEC1 TSEC2 SEC2 AL SEC2 STI M NOTE IDNT ANOTE RIDNT SEC1 TSEC1SEC2 TSEC2 AL TSEC2 AL NOTE D'ETUDE 2EME STADE E M E LM 99 0102 Pièce N° 1 Note de synthèse 1.1.1 Historique 1.1 origine de la modification 1.1.2 Énoncé du problème - But recherché La note d'étude E-XX/FC 2.17 ind. A attire l'attention sur le fait que le temps de blahblah important dû à la complexité du système de verrouillage est paradoxal pour un système de secours. Faisant suite à cette conclusion le SEPTEN (note d'étude E-XX/FC 2.18) ind. A que blahblah… 1.2 Description de la modification 1.2.1 Solution retenue Le XXX des YYY se traduit par la mise en place de ZZZ dans les armoires d'aiguillage permettant de ligner le ABC vers la totalité des tableaux RRR du site 1 CCE 3.14159 du 31/12/03 La logique de verrouillage associée au système ABC a pour but de garantir la sécurité du personnel, lors des manœuvres d’exploitation. blahblah Depuis sa création, les objectifs en matière de sûreté et de disponibilité du système ABC ont évolué. A ce titre la CCE 3.14159 1 a confié au SEPTEN une étude précisant pour les différents états des tranches les conditions d'utilisation des BZZ. A la suite de l'étude SEPTEN le CTE du 30/02/1995 donne son accord pour la réalisation du xxx de yyy sur zzz A document is a « Tree of typed objects »
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 12 XTM URIUnicode Namespaces RDF XML DTDsXML Schémas HTML “Be fertile and multiply; fill the earth and subdue it”. So did websites, they multiply and invade the globe… Searching and finding published information, build indexes of access to documents became a challenge –To which web search engines provided solutions But with plenty of noise and silence A structured meaningful index, building a concept network cannot be expressed as a simple tree –Topic Maps (ISO Standard) et RDF(W3Cstandard) provide two ways of representation of networks of concepts using the XML language
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 13 The fence problem ? pickets or intervals, that is the question ! fences in a country may be seen as networks of pickets linked together, or as barriers intervals connected through pickets RDF and Topic Maps differs by their approach – linguists used to prefer a “picket” approach – Artificial Intelligence used to prefer the “interval” approach XTM Topic Maps RDF Resource Description Framework
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 14 About Topic Maps From Wikipedia : –Topic Maps is an ISO standard for the representation and interchange of knowledge, with an emphasis on the ability to find information. The standard is formally known as ISO/IEC 13250:2003. –A topic map can represent information using topics (representing any concept, from people, countries, and organizations to software modules, individual files, and events), associations (which represent the relationships between them), and occurrences (which represent relationships between topics and information resources relevant to them). –Topic Maps are thus similar to semantic networks and both concept and mind maps in many respects. In loose usage all those concepts are often used synonymously, though only topic maps are standardized.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 15 Example of Index using Topic Maps : a document content navigation system
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 16 XTM Two kinds of complementary model descriptions URIUnicode Namespaces RDF RDF Schema Ontology XML DTDsXML Schémas HTML XML Schema and DTDs for the description of document structure RDFSchema et OWL for semantic description of knowledge networks
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 17 Ontologies… …formally expressed Ontology is a study of conceptions of reality and the nature of being.reality In philosophy, ontology (from the Greek ὤ ν, genitive ὄ ντος: of being (part. of ε ἶ ναι: to be) and -λογία: science, study, theory) is the study of being or existence and forms the basic subject matter of metaphysics.-λογία –It seeks to describe or posit the basic categories and relationships of being or existence to define entities and types of entities within its framework. In both computer science and information science, an ontology is a data model that represents a set of concepts within a domain and the relationships between those concepts. It is used to reason about the objects within that domain. (…that is an other way to express the same thing) –an ontology is a structure set of concepts. –Concepts are organized as a graph in which relations can be : semantic relations ; composition relations and inheritance relations (au sens objet) –Knowledge that refer one or several ontologies use their vocabularies without ambiguity Ontology … From Parmenides to Tim Berners Lee
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 18 UML XMI Where may be situated UML … URIUnicode Namespaces RDF RDF Schema Ontology XML DTDsXML Schémas HTML Any UML diagram can be expressed as a populated ontology But XMI has been defined as a low level XML idiom…
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 19 Expressing Rules Rules Logic Framework SWRL (Semantic Web Rule Language) is a proposal for a Semantic Web rules-language, combining sublanguages of the OWL Web Ontology Language (OWL DL and Lite) with those of the Rule Markup Language (Unary/Binary Datalog). The specification was submitted in May 2004 to the W3C.2004 The Rule Interchange Format (RIF) is a W3C recommendation-track effort to develop a format for interchange of rules in rule-based systems on the semantic web. The goal is to create an interchange format for different rule languages and inference engines. The RIF initiative is closely related to Ontologies. Whereas ontologies describe distributed information objects in a computer executable manner, rules in this sense combine such information and derive new information on top of ontologies. (source Wikipedia)
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 20 Populated ontology gets benefits from the logic of their description URIUnicode Namespaces RDF RDF Schema Ontology Rules Logic Framework Encryption Signature XML DTDsXML Schémas HTML Linked to inference engines, ontologies gets benefits from artificial intelligence reasoning on knowledge
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 21 XTM UML XMI Sir Tim Berners Lee’s full vision, URIUnicode Namespaces RDF RDF Schema Ontology Rules Logic Framework Proof Trust Encryption Signature XML DTDsXML Schémas HTML
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 22 RDF Resource Description Framework
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 23 Issues of the access to Web resources «World Wide Web was initially designed for human understanding, and while all that is published on the web is readable by a computer, these data are not understood by computers. » « It is hard to build automatic processing of data on the web, but due to the amount of published data, it is impossible to manage it manually ». « The proposed solution is to use metadata in order to describe data published on the Web ». « Metadata are data about data ( for instance a library catalogue is a compilation of metadata as it is a description of the books its contain) ; in the web case metadata are data about web resources. » « Difference between data and metadata is not obvious. It depends on specific needs and on applications that fulfill these needs. A same resource may be tackled simultaneously in different ways »
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 24 Le Dublin Core est un ensemble de 15 éléments de métadonnées ayant trait: –au Contenu: Title, Description, Subject, Source, Coverage, Type, Relation –à la Propriété intellectuelle: Creator, Contributor, Publisher, Rights –à la Version: Date, Format, Identifier, Language Les balises du Dublin core forment un vocabulaire non ambigu par leur association à l’espace de nom du Dublin Core –http://purl.org/dc/elements/1.1/ Information about rights held in and over the resourceRightsdroits The extent or scope of the content of the resource including spatial location temporal period or jurisdictionCoveragecouverture A reference to a related resource. Recommended best practice is to reference the resource by means of a string or number conforming to a formal identification system. Relationrelation A language of the intellectual content of the resourceLanguagelangue A Reference to a resource from which the present resource is derivedSourcesource An unambiguous reference to the resource within a given contextIdentifieridentifiant The physical or digital manifestation of the resource.Formatformat The nature or genre of the content of the resource (Collection Dataset Event Image InteractiveResource MovingImage PhysicalObject Service Software Sound StillImage Tex)t Typetype A date linked to an event in the lifecycle of the resourceDatedate Any entity, person or organization that contributes to the content of the resourceContributorcontributeur Organization responsible of the publication of the resource, may be fo instance an university department, or a company.Publisheréditeur description of the resource contentDescriptiondescription Subject of the resource contentSubjectsujet et mots-clefs Organization or person responsible of the resource creationCreatorcréateur Resource nameTitletitre DefinitionIdentifierFrench name The « Dublin Core » initiative
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 25 RDF : an introduction RDF (Resource Description Framework) is a mean of register, exchange and reuse structured metadata in order ti describe data. It is an XML idiom developed by W3C and of which the first recommendation was published in 1999. RDF do not give a set of fixed semantic of the resources described by different users communities. Euach of them may introduce its own vocabulary using a specific namespace and a schema giving a descrption of each term used. Based on XML, RDF is also an extensible language i.e. a meta language il is a framework of resource description tat may be applicable to any application domain.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 26 RDF triples RDF expressions are triples –Subject predicate object ou –Resource property value Exemple : –The document XML Schéma was written by Eric Van der Vlist subjet predicate object These triples are formalized as oriented graphs XML Schema Author predicate subject (resource) object (value) Eric Van der Vlist
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 27 Resource Identification Resources are identified through URI (Unified Resource Identifier). URI are set of names used to address any thing or concept (URL Unified Resource Locator are kind of URI). In our example the document XML Schema may be identified by its URI: http://www.oreilly.fr/XML_Schema.pdf predicates (properties) are also identified using URI –The URI of the predicate “author" may so be the Creator element of the Dublin Core schema: http://purl.org/dc/elements/1.1/creator
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 28 « Bootstrap » of the RDF Syntax RDF is designed to be a description language of facts and of relationships between these facts. Even if that is the expression of a graph, it is a fact that any graph may be supported by a tree structure. That allows the expression of RDF using the XML syntax principles, using the root element : As the aim of RDF is facts description, the first element of RDF expressions is « Description » A description is like a sequenced story attached to facts introduced using an identifier, and then referred A description links fact with properties, of which the meaning may by known only by their association to a defined vocabulary. Such as the Dublin Core vocabulary : xmlns:dc="http://purl.org/dc/elements/1.1/" a fact is expressed through the association of a value to a property of anything we want to describe: un titre A property may have a literal content or refer an other fact, which need also… … to be described So is the XML Syntax principle for an RDF triple Subject / Predicate /Object
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 29 Groups, categories, classes, and all that sort of things So is a common characteristic of human languages, perhaps due to the architecture of their brain, in order to remember facts and speak of them the need to categorize them ! So are people uses, in order to simplify their expressions they designates facts using category names And give name of characteristics as name of these categories!
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 30 EBNF Formalization (Extended Backus-Naur form) of RDF grammar  RDF ::= [' '] description* [' ']  description ::= ' ' propertyElt* ' '  idAboutAttr ::= idAttr | aboutAttr  aboutAttr ::= 'about="' URI-reference '"'  idAttr ::= 'ID="' IDsymbol '"'  propertyElt ::= ' ' value ' ' | ' '  propName ::= Qname  value ::= description | string  resourceAttr ::= 'resource="' URI-reference '"‘  Qname ::= [ NSprefix ':' ] name  URI-reference ::= string, interpreted per [URI]URI  IDsymbol ::= (any legal XML name symbol)XML name symbol  name ::= (any legal XML name symbol)  NSprefix ::= (any legal XML namespace prefix)XML namespace prefix  string ::= (any XML text, with " ", and "&" escaped) [2a] description ::= ' ' | ' ' propertyElt* ' ' | typedNode [6a] propertyElt ::= ' ' value ' ' | ' '  propAttr ::= propName '="' string '"' (with embedded quotes escaped)  typedNode ::= ' ' | ' ' property* ' '
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 31 XML Syntax of RDF (cont’d) a subject (resource) may own several predicates (properties) That is expressed in the RDF syntax : Eric Van der Vlist URL of Dublin Core predicates Subject Predicate object
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 32 RDF : graphs expression Les resources may be linked and embedded: xmlns:s="http://description.org/schema/"> Eric Van der Vlist VDV@XMLfr.org XML Schema XMLfr.org creator name mail VDV@XMLfr.org Eric Van der Vlist
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 33 Basic terms of the RDF Vocabulary RDF vocabulary namespace http://www.w3.org/1999/02/22-rdf-syntax-ns –Resource : Any “thing” refered by an RDF expression, identified by an URI –Property : Any aspect, characteristic of a Resource, its relation with another Resource ; Is itself a kind of Resource –Statement : expression of the value of a named property attached to a given Resource. Is a triple “subject / predicate / value” reification
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 34 Schemas and Namespaces What is required for the human understanding of statements is also required for computer applications using RDF in order to get proper results. It is required that writer and reader of a statement get the same understanding of the used terms meaning, such as Creator, ApprovedBy, Copyright. Any confusion, or mismatch of concepts understanding should be avoided In RDF the meaning is expressed through reference to a schema A schema is a kind of dictionary. It defines terms to be used in RDF statements and give them unique meanings Different forms of schema may be used in RDF, including a specific one defined in a separate document [RDFSchema] which has a set of specific characteristics defined in order to help task automation using RDFRDFSchema
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 35 Culminate : two different concepts for two different aims One should make difference between: –Schema XML that express structure constraints and syntax for XML “documents” –Schema RDF that express semantic constraints building model of RDF “descriptions” Concept of RDF schema –RDF schema describe a vocabulary and a semantic for the property types utilized by a community of users. –RDF schema define the valid properties foe a specific RDF description, as well as the characteristics and constraints of the description vocabulary. –As an example on may look at the Dublin Core RDF Schema version 1.1
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 36 RDF Schema A schema is the place where properties definitions and usage restrictions are documented. In order to avoid confusion between independents definitions – with possible conflicts – of a same term, RDF implements the XML namespaces features. Namespaces are just a way to link any specific use of a term in a specific context to the dictionary (schema) where its definition is supposed to be defined. Each RDF predicate used in a statement must be identified with exactly one namespace or a schema. A Description element may contain statements with predicates from several different de schemas.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 37 Abbreviated syntax Full Description : Renault abbreviated syntax, about types : Other kind of abbreviated syntax about attributes :
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 38 Containers It is often necessary to refer a collection of resources. For instance a group of people, a list of actions to do, or to choose. RDF containers are designed to express such lists of resources or literals. RDF define three kinds of containing objects : –Bag a list of resources or literals without order. Bag is used to set that a property owns several values and that there is no meaning attached to the way they are listed. Identical values are allowed. –Sequence an ordered sequence of resources or literals. The order of values of a property is heare meaningful. Identical values are allowed. –Alternative A list of resources or literals that represent alternative choices for the unique value of a property.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 39 Examples CD Record Tape George John Paul Ringo
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 40 Closed Collections The attribute rdf:parseType="Collection" Define a non ordered closed list of elements #
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 41 N3 : An other kind of RDF expression N3 notation is simple : subjectverbobjectpunctuation. (All is URL ; the character # indicate that the resource is attached to the URI of the current document courant, which namespace is a default ) Only object can be literal : 24. More readable : is 24. punctuation ";" et"," allows to factorize the subject respectively for verb and object is of ; is of. [ ] identify resource from properties has [ 21], [ 24], Other kind of expression : [ "Pat"; 24; "blue" ].
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 42 N3 : An other kind of RDF expression (cont’d) The current document is sayed as followed <> "A simple example of N3". However attachment to a namespace is preferable : <> " titre de ce document ". prefix mechanism used to refer a namespace : @prefix dc:. <> dc:title " titre de ce document ". (where the fact that the name is colonized allow to avoid de <>) To declare the use of a default prefix : @prefix :. Then one may write : :pat :aged 24. au lieu de 24. vocabulary equivalence is expressed using the sign = :woman = foo:human_female. :Title a rdf:Property; = dc:title.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 43 Exercise Description in RDF of the relationships in the town of Verona between Montaigu and Capulets, Roméo and Juliette… Vérone Montaigu Capulet RoméoJuliettelove hate is Live in
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 44
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 45 RDFS First Initiative of description of RDF resources organisation Extension of RDF, as RDF define a basic set of classes and properties. In the beginning was… …the « resource » from which all come… –rdfs:Resource : class of which all class is a sub-type –rdfs:Class, instance of itself, sub-type of rdfs:Resource –rdfs:Literal instance of rdfs:Class, sub-type of rdfs:Resource –rdfs:Datatype is the class of attributes. its instances correspond to a data model described in the specification of RDF concepts. rdfs:Datatype is as well an instance and a sub-class of rdfs:Class. Each instance of rdfs:Datatype is a sub-type of rdfs:Literal. –rdfs:Property it the class of RDF properties, instance of rdf:Class –rdfs:range and rdfs:domain are instances of rdfs:Property attached recursively to rdfs:Property, : the domain rdfs:domain de rdfs:range and of rdfs:domain is rdfs:Property… –rdf:type is an instance of rdfs:Property used to define a resource as an instance of a class.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 46 RDFS (suite) –rdfs:subClassOf is an instance of rdfs:Property which defines that instances of a classe are instances of an other classe from which its inherits –rdfs:subPropertyOf is an instance of rdfs:Property which defines that all resources addressed by a property are also resources of an other property from which its inherits. –rdfs:label is an instance of rdfs:Property used to give a human readable version of a resource. –rdfs:comment is an instance of rdfs:Property used to document a resource by a comment
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 47 The class of RDF Lists.rdf:List The class of container membership properties, rdf:_1, rdf:_2,..., all of which are sub-properties of 'member'. rdfs:ContainerMembershipProperty The class of RDF containers.rdfs:Container The class of containers of alternatives.rdf:Alt The class of ordered containers.rdf:Seq The class of unordered containers.rdf:Bag The class of RDF statements.rdf:Statement The class of RDF datatypes.rdfs:Datatype The class of RDF properties.rdf:Property The class of classes.rdfs:Class The class of XML literals values.rdf:XMLLiteral The class of literal values, e.g. textual strings and integers.rdfs:Literal The class resource, everything.rdfs:Resource commentClass name
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 48 rdfs:Resourcerdf:StatementThe object of the subject RDF statement.rdf:object rdfs:Resourcerdf:StatementThe predicate of the subject RDF statement.rdf:predicate rdfs:Resourcerdf:StatementThe subject of the subject RDF statement.rdf:subject rdfs:Resource Idiomatic property used for structured values (see the RDF Primer for an example of its usage).an example rdf:value rdfs:Resource The definition of the subject resource.rdfs:isDefinedBy rdfs:Resource Further information about the subject resource.rdfs:seeAlso rdf:List The rest of the subject RDF list after the first item.rdf:rest rdfs:Resourcerdf:ListThe first item in the subject RDF list.rdf:first rdfs:Resource A member of the subject resource.rdfs:member rdfs:Literalrdfs:ResourceA description of the subject resource.rdfs:comment rdfs:Literalrdfs:ResourceA human-readable name for the subject.rdfs:label rdfs:Classrdf:PropertyA range of the subject property.rdfs:range rdfs:Classrdf:PropertyA domain of the subject property.rdfs:domain rdf:Property The subject is a subproperty of a property.rdfs:subPropertyOf rdfs:Class The subject is a subclass of a class.rdfs:subClassOf rdfs:Classrdfs:ResourceThe subject is an instance of a class.rdf:type rangedomaincommentProperty name
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 49 Tragedy Sample
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 50 OWL OWL : acronym of Web Ontology Language (en plus chouette!) –Wikipedia : The Web Ontology Language (OWL) is a language for defining and instantiating Web ontologies. An OWL ontology may include descriptions of classes, along with their related properties and instances. OWL is designed for use by applications that need to process the content of information instead of just presenting information to humans. It facilitates greater machine interpretability of Web content than that supported by XML, RDF, and RDF Schema (RDF-S) by providing additional vocabulary along with a formal semantics. OWL is based on earlier languages OIL and DAML+OIL, and is now a W3C recommendation..WebontologiesXMLRDFRDF SchemasemanticsOILDAML+OIL W3Crecommendation –To the concepts of class, resource, literal and properties of sub-classes, de sub- properties, of value fields, of application domains already existing in RDFS, OWL add concepts of equivalent classes, of equivalent properties, equivalent resources, different resources, opposite, de symmetric resources. It adds also the concept of cardinality... The OWL documents became a formal W3C recommendation on February 10, 2004 and the working group was disbanded on May 31, 2004.February 10 2004May 312004
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 51 The three levels of OWL Wikipedia : OWL currently has three sublanguages (sometimes also referred to as 'species'): OWL Lite, OWL DL, and OWL Full. These three increasingly expressive sublanguages are designed for use by specific communities of implementers and users. : –OWL-Lite,OWL-Lite supports those users primarily needing a classification hierarchy and simple constraints. For example, while it supports cardinality constraints, it only permits cardinality values of 0 or 1. It should be simpler to provide tool support for OWL Lite than its more expressive relatives, and OWL Lite provides a quick migration path for thesauri and other taxonomies. OWL Lite also has a lower formal complexity than OWL DL; see the section on OWL Lite in the OWL Reference for further details.cardinalitythesauri taxonomies –OWL-DL,OWL-DL supports those users who want the maximum expressiveness while retaining computational completeness (all conclusions are guaranteed to be computed) and decidability (all computations will finish in finite time). OWL DL includes all OWL language constructs, but they can be used only under certain restrictions (for example, while a class may be a subclass of many classes, a class cannot be an instance of another class). OWL DL is so named due to its correspondence with description logic, a field of research that has studied the logics that form the formal foundation of OWLdecidabilitydescription logic –OWL-Full.,OWL-Full is meant for users who want maximum expressiveness and the syntactic freedom of RDF with no computational guarantees. For example, in OWL Full a class can be treated simultaneously as a collection of individuals and as an individual in its own right. OWL Full allows an ontology to augment the meaning of the pre-defined (RDF or OWL) vocabulary. It is unlikely that any reasoning software will be able to support complete reasoning for every feature of OWL Full.
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 52 OWL Class Owl:Thing is the mother class of all classes Owl:noThing is the child class of all classes An OWL class may be defined – by a reference (URI) – through a list of instances:
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 53 OWL Class An OWL class may be defined – by its properties (intension definition) Properties Types : owl:allValuesFrom,owl:someValuesFrom Properties Values : owl:hasValue Properties Cardinality : owl:maxCardinality, owl:minCardinality, owl:Cardinality Example : class of resources which have one or more property #author of type #novelist
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 54 OWL Class An OWL class may be defined –as union, intersection, complement of other classes
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 55 rdfs:subClassOf : the extension of a class is included in the extension of the other owl:equivalentClass : both classes have same extension, but are not of the same concept. owl:disjointWith : disjoint classes OWL Class
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 56 Properties RDF Schema provide the concepts of: –rdfs:subPropertyOf, rdfs:domain and rdfs:range OWL enlarge relationship between properties : –owl:equivalentProperty : both have same extension, but are not identicals : –owl:inverseOf : one is the inverse of the other. –cardinality constraints de : owl:FunctionalProperty (mono-valuées), et owl:InverseFunctionalProperty –logical constraints : owl:SymmetricProperty (exemple époux) owl:TransitiveProperty (exemple ascendant)
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 57 Shakespeare example (ontology)
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 58 Shakespeare example (ontology)
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 59 Shakespeare example (ontology)
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 60 Shakespeare example (populated ontology) Roméo Vérone Juliette Capulet
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 61 Semantic Web Rule Language (swrl) SWRL (Semantic Web Rule Language) is a rule language for the semantic web, that combine des sub-languages OWL (OWL DL and OWL Lite) with those of Rule Markup Language (Unary/Binary Datalog). Exemple of family rule: hasParent(?x1,?x2) ∧ hasBrother(?x2,?x3) ⇒ hasUncle(?x1,?x3) x1 x2 x3 x1 x3
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 62 Huge set of syntax… … but efficient tools exist that allow to build and populate ontologies, for instance : –Protégé From Stanford University, is the most known, popular and used ontology tool. It is an Open-source software, initially based on a frame model, adapted to RDF, RDFS OWL standards. It is a platform on which plug-ins may be connected for reasoning, graphical display etc. http://protege.stanford.edu/ –Swoop Is an ontology development tool created by the Maryland University for the MINDSWAP project. It has been natively based on RDF and OWL standards, with their different syntaxes (not only XML). It is a lighter application than Protégé but it accepts the connection of inference engines. http://www.mindswap.org/2004/SWOOP/
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 63 Culminate 2 : Yet an other standard ! For what is it useful ? As for XML, plenty of RDF/RDFS/OWL applications do not initially depend on the Web. –Ontologies are formal standard expressions of kind of data managed by popular mind mapping tools Mind manager The Brain, Visual Mind SmartDraw FreeMind OWL Ontologies do not have the rigidity of SQL database models –One can expand and improve easily models of populated ontology, without loss of data
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 64 Development of ontologies contents Plenty ontologies are already used to store huge knowledge –Generic knowledge OpenCyc ; Cyc is an artificial intelligence project that attempts to assemble a comprehensive ontology and database of everyday common sense knowledge, with the goal of enabling AI applications to perform human-like reasoning., Sumo, Suggested Upper Merged Ontology, is an upper ontology intended as a foundation ontology for a variety of computer information processing systems. It was originally developed by the Teknowledge Corporation and now is maintained by Articulate Software. It is one candidate for the "standard upper ontology" that IEEE working group 1600.1 is working on. It can be downloaded and used freely. –A Generic Ontology Search engine http://swoogle.umbc.edu/ –Knowledge In biological sciences Gene ontology –http://geneontology.org/index.shtmlhttp://geneontology.org/index.shtml –http://archive.geneontology.org/latest-termdb/go_daily-termdb.owl.gzhttp://archive.geneontology.org/latest-termdb/go_daily-termdb.owl.gz Open Biomedical Ontologies –OBO, Another ontological expression : http://genomebiology.com/2005/6/5/R46 http://genomebiology.com/2005/6/5/R46 –http://www.berkeleybop.org/ontologies/#ontologieshttp://www.berkeleybop.org/ontologies/#ontologies –…–…
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 65 Lexter® Exter® Worldtrek® XTM Topic Maps publication Web Worldtrek® Human Expertise corpus TML Term candidates network TML Structured Validated Terminology Human knowledge texts resources TMF concepts & terms corpus OWL Ontology Classes Individuals Terminology A pragmatic ontology building method: using a linguistic terminology to ontology chain
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 66 A pragmatic ontology building method: using a linguistic terminology to ontology chain OWL-DL rules SWRL rules knowledge Application Web portal corpus TMF OWL OWL editor Protégé® Swoop® … TML20WL (XSL) OWL2TML Jena CleanOWL® (Java) accessed classes description prescription operation inference other services
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 67 Example of ontology construction using the WorldTrek terminology edition tool
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 68 Result of terminology edition in : Protégé®
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 69 Pragmatic ontologies application : for Services Oriented Architecture (SOA) Ontologies are pragmatic ways for web service discovery –A semantic network index is a better solution than a flat directory –OWL-S initiative provides mechanism to manage Web services metadata Service Grounding Service Profile Service Model Has description presentsupport What services are proposed ? How services work ? How to access to services ?
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 70 Pragmatic ontologies application : Enterprise Information System architecture
ATHENS 2007 - ENSAM Paris Henry Boccon-Gibod 71 Perspective, in order to conclude ” People keep asking what Web 3.0 is. I think maybe when you've got an overlay of scalable vector graphics - everything rippling and folding and looking misty - on Web 2.0 and access to a semantic Web integrated across a huge space of data, you'll have access to an unbelievable data resource. ” Tim Berners-Lee,