The Semantic Web

1

• The Semantic Web
• and (vs?)
• Knowledge Representation
• ENC 2004, September 2004

Peter F. Patel-Schneider Bell Labs
Research Murray Hill, NJ, USA pfps_at_lucent.com
2
Abstract

• The vision of the Semantic Web is a collection of
  documents in the World Wide Web whose content is
  meaningful to computers. In this vision
  computers can directly and effectively process
  information stored in the Web without having to
  depend on human guidance. This vision also
  underlies Knowledge Representation, a
  subdiscipline of Artificial Intelligence.
  Research in Knowledge Representation over the
  last four decades has produced some surprising
  results but has also identified some very
  difficult problems that need to be overcome if
  this vision is to be realized in its entirety.
  The Semantic Web vision provides several
  opportunities for Knowlege Representation that
  have not been explioted in the past as well as
  some new difficulties that could hinder
  application of techniques from Knowledge
  Representation to the Semantic Web. Achieving
  the vision of the Semantic Web require exploiting
  the successes of Knoweldge Representation in a
  challenging new environment.

3
Acknowledgments and Caveat

• Some of the slides for this talk have been taken
  from various talks on the Semantic Web by Ian
  Horrocks and others, including Jeen Broekstra,
  Carole Goble, Frank van Harmelen, Austin Tate,
  and Raphael Volz. Thanks go to all of them.
• This talk contains my personal opinions on the
  Semantic Web and Knowledge Representation. Other
  researchers in both areas have differing opinions.

4
What is the Semantic Web?

• The Semantic Web aims to make information on the
  World Wide Web accessible to computers.
• Not only parsable by computers (i.e., XML), but
  also understandable (in some sense) by computers.
• Prior agreements between humans are not needed to
  provide meaning (as is the case for XML).
• Human guidance is not always needed.

5
History of the Semantic Web

• Web was invented by Tim Berners-Lee (amongst
  others), a physicist working at CERN
• TBLs original vision of the Web was much more
  ambitious than the reality of the existing
  (syntactic) Web

• TBL (and others) have since been working towards
  realising this vision, which has become known as
  the Semantic Web
• E.g., article in May 2001 issue of Scientific
  American

6
Scientific American, May 2001
7
Scientific American, May 2001

• Realising the complete vision is too hard for
  now (probably)
• But we can make a start by adding semantic
  annotation to web resources

8
What is Knowledge Representation?

• Knowledge Representation aims to make information
  accessible to computers.
• Not only parsable by computers (i.e., databases)
  but also understandable (in some sense) by
  computers.
• Prior agreements between humans are not needed to
  provide meaning (as is the case for databases).
• Human guidance is not always needed.

9
Why the Semantic Web matches Knowledge
Representation

• Semantic Web is not just data
• Divergent interpretations of reality
• Missing information (and not just null values)
• Doesnt match assumptions of databases or
  object-oriented programming
• Open world (no closed-world assumption)
• Objects change status over time
• Semantic Web is a representation system

10
Differences between Knowledge Representation and
Databases

• Database (relational, object-oriented,
  semi-structured) assumptions
• All relevant information is known (not there
  implies not true)
• All information is definite
• No disjunction, as in Johns friend is either
  Susan or Bill
• Objects have a single minimal class/type
• Cant have John is both a student and an employee
  (unless there is a student-employee class/type)
• Knowledge Representation assumptions
• Relevant information may be missing
• E.g., no information about Johns friends doesnt
  mean that he doesnt have any
• Indefinite information may be present
• E.g., John has a friend, but who it is is not
  known
• Multiple (and changing types) allowed

11
The Semantic Web and/vsKnowledge Representation

• The Semantic Web is both an opportunity and a
  challenge for Knowledge Representation
• An opportunity because the Semantic Web is (or
  will be) a source of information with very
  similar goals to those of Knowledge
  Representation
• A challenge because some of the characteristics
  of the Semantic Web violate some of the
  assumptions that have generally been held in
  Knowledge Representation
• Knowledge Representation is both a resource and a
  cautionary tale for the Semantic Web
• Knowledge Representation techniques can be
  utilized in the Semantic Web
• Problems encountered in Knowledge Representation
  have already plauged the Semantic Web.

12
Why Knowledge Representation is a resource for
the Semantic Web

• Knowledge Representation provides formal rigor
• Meaning of information-bearing constructs are
  formally determined
• Needed for computers to process the constructs
  (compare with formal syntax, also required for
  computer processing)
• Knowledge Representation is concerned with
  reasoning
• Determining what follows from a collection of
  information
• Provides an account of what can be (reliably)
  done by a computer
• Knowledge Representation systems are becoming
  quite powerful and reliable

13
Why Knowledge Representation is a cautionary tale
for the Semantic Web

• Early Knowledge Representation was not formal
• Lead to many interminable debates about the
  meaning of constructs
• Current work in Knowledge Representation is very
  concerned with the formal meaning of constructs
• Computing with representations is difficult
• Many problems are intractable or even undecidable
• Lead to a retreat to simpler languages
• Current systems are quite capable, even on
  expressive languages
• Heavily optimized code
• Computers are much more powerful

14
Why the Semantic Web is an opportunity for
Knowledge Representation

• The World Wide Web forms a (very) large source of
  information (albeit in awkward formats)
• The Semantic Web aims to transform much of this
  information into a form compatible with the goals
  of Knowledge Representation
• The Semantic Web also will contain services that
  can be controlled by other computers (and
  reasoned about)
• A potential solution to the Grounding Problem in
  Knowledge Representation

15
Why the Semantic Web is a challenge for Knowledge
Representation

• Very large amounts of information will be part of
  the Semantic Web
• Can overwhelm formal reasoning methods
• The Semantic Web contains differing
  interpretations of reality
• Which one(s) to choose?
• Recovering from inconsistencies
• How to determine how inconsistencies arise
• How to determine who to trust

16

• Why is the Semantic Web
• a good idea?

17
Where we are Today the Syntactic Web
Hendler Miller 02
18
The Syntactic Web is

• A hypermedia, a digital library
• A library of documents called (web pages)
  interconnected by a hypermedia of links
• A database, an application platform
• A common portal to applications accessible
  through web pages, and presenting their results
  as web pages
• A platform for multimedia
• BBC Radio 4 anywhere in the world! Terminator 3
  trailers!
• A naming scheme
• Unique identity for those documents
• A place where computers do the presentation
  (easy) and people do the linking and interpreting
  (hard).
• Why not get computers to do more of the hard
  work?

Goble 03
19
Hard Work using the Syntactic Web
Find images of Peter Patel-Schneider, Frank van
Harmelen and Alan Rector
Rev. Alan M. Gates, Associate Rector of the
Church of the Holy Spirit, Lake Forest, Illinois
20
Impossible (?) using the Syntactic Web

• Complex queries involving background knowledge
• Find information about animals that use sonar
  but are not either bats or dolphins
• Locating information in data repositories
• Travel enquiries
• Prices of goods and services
• Results of human genome experiments
• Finding and using web services
• Visualise surface interactions between two
  proteins
• Delegating complex tasks to web agents
• Book me a holiday next weekend somewhere warm,
  not too far away, and where they speak French or
  English

21
What is the Problem?

• Consider a typical web page

• Markup consists of
• rendering information (e.g., font size and
  colour)
• Hyper-links to related content
• Semantic content is accessible to humans but not
  (easily) to computers

22
What information can we see

• WWW2002
• The eleventh international world wide web
  conference
• Sheraton waikiki hotel
• Honolulu, hawaii, USA
• 7-11 may 2002
• 1 location 5 days learn interact
• Registered participants coming from
• australia, canada, chile denmark, france,
  germany, ghana, hong kong, india, ireland, italy,
  japan, malta, new zealand, the netherlands,
  norway, singapore, switzerland, the united
  kingdom, the united states, vietnam, zaire
• Register now
• On the 7th May Honolulu will provide the backdrop
  of the eleventh international world wide web
  conference. This prestigious event
• Speakers confirmed
• Tim berners-lee
• Tim is the well known inventor of the Web,
• Ian Foster
• Ian is the pioneer of the Grid, the next
  generation internet

23
What information can a machine see

• WWW2002
• The eleventh international world wide web
  conference
• Sheraton waikiki hotel
• Honolulu, hawaii, USA
• 7-11 may 2002
• 1 location 5 days learn interact
• Registered participants coming from
• australia, canada, chile denmark, france,
  germany, ghana, hong kong, india, ireland, italy,
  japan, malta, new zealand, the netherlands,
  norway, singapore, switzerland, the united
  kingdom, the united states, vietnam, zaire
• Register now
• On the 7th May Honolulu will provide the backdrop
  of the eleventh international world wide web
  conference. This prestigious event
• Speakers confirmed
• Tim berners-lee
• Tim is the well known inventor of the Web,
• Ian Foster
• Ian is the pioneer of the Grid, the next
  generation internet

24
Solution XML markup with meaningful tags?

• ltnamegtWWW2002
• The eleventh international world wide
  webconlt/namegt
• ltlocationgtSheraton waikiki hotel
• Honolulu, hawaii, USAlt/locationgt
• ltdategt7-11 may 2002lt/dategt
• ltslogangt1 location 5 days learn interactlt/slogangt
• ltparticipantsgtRegistered participants coming from
• australia, canada, chile denmark, france,
  germany, ghana, hong kong, india, ireland, italy,
  japan, malta, new zealand, the netherlands,
  norway, singapore, switzerland, the united
  kingdom, the united states, vietnam,
  zairelt/participantsgt
• ltintroductiongtRegister now
• On the 7th May Honolulu will provide the backdrop
  of the eleventh international world wide web
  conference. This prestigious event
• Speakers confirmedlt/introductiongt
• ltspeakergtTim berners-leelt/speakergt
• ltbiogtTim is the well known inventor of the
  Web,lt/biogt

25
But What About

• ltconfgtWWW2002
• The eleventh international world wide
  webconlt/confgt
• ltplacegtSheraton waikiki hotel
• Honolulu, hawaii, USAlt/placegt
• ltdategt7-11 may 2002lt/dategt
• ltslogangt1 location 5 days learn interactlt/slogangt
• ltparticipantsgtRegistered participants coming from
• australia, canada, chile denmark, france,
  germany, ghana, hong kong, india, ireland, italy,
  japan, malta, new zealand, the netherlands,
  norway, singapore, switzerland, the united
  kingdom, the united states, vietnam,
  zairelt/participantsgt
• ltintroductiongtRegister now
• On the 7th May Honolulu will provide the backdrop
  of the eleventh international world wide web
  conference. This prestigious event
• Speakers confirmedlt/introductiongt
• ltspeakergtTim berners-leelt/speakergt
• ltbiogtTim is the well known inventor of the Web,

26
Machine sees

• ltnamegtWWW2002
• The eleventh international world wide webclt/namegt
• ltlocationgtSheraton waikiki hotel
• Honolulu, hawaii, USAlt/locationgt
• ltdategt7-11 may 2002lt/dategt
• ltslogangt1 location 5 days learn interactlt/slogangt
• ltparticipantsgtRegistered participants coming from
• australia, canada, chile denmark, france,
  germany, ghana, hong kong, india, ireland, italy,
  japan, malta, new zealand, the netherlands,
  norway, singapore, switzerland, the united
  kingdom, the united states, vietnam,
  zairelt/participantsgt
• ltintroductiongtRegister now
• On the 7th May Honolulu will provide the backdrop
  of the eleventh international world wide web
  conference. This prestigious event
• Speakers confirmedlt/introductiongt
• ltspeakergtTim berners-leelt/speakergt
• ltbiogtTim is the well known inventor of the
  Wlt/biogt
• ltspeakergtIan Fosterlt/speakergt
• ltbiogtIan is the pioneer of the Grid, the nelt/biogt

27
Need to Add Semantics

• Two very different possible approaches
• External agreement on meaning of annotations
• Agree on the meaning of a set of annotation tags,
  e.g., Dublin core
• Problems with this approach
• Inflexible
• Limited number of things can be expressed
• Use on-line Ontologies to specify meaning of
  annotations
• Ontologies provide a vocabulary of terms
• New terms can be formed by combining existing
  ones
• Meaning (semantics) of such terms is formally
  specified
• Can also specify relationships between terms in
  multiple ontologies
• Semantic Web takes second approach

28
Characteristics of the Semantic Web

• Part of the Web
• Uses Web addressing (URIs)
• Adheres to Web philosophy
• Connected to the rest of the Web (e.g, services)
• Large part of the Web
• Very many connected documents
• Diverse, conflicting
• Semantic
• Contains ontological information about meaning of
  objects

29

• What are
• Ontologies?

30
Ontology Origins and History
Ontology in Philosophy

• a philosophical disciplinea branch of
  philosophy that
• deals with the nature and the organisation of
  reality
• Science of Being (Aristotle, Metaphysics, IV, 1)
• Tries to answer the questions
• What characterizes being?
• Eventually, what is being?

31
Ontology in Linguistics
Tank
32
Ontology in Computer Science

• An ontology is an engineering artifact
• It is constituted by a specific vocabulary used
  to describe a certain reality, plus
• a set of explicit assumptions regarding the
  intended meaning of the vocabulary.
• Thus, an ontology describes a formal
  specification of a certain domain
• Shared understanding of a domain of interest
• Formal and machine manipulable model of a domain
  of interest
• An explicit specification of a
  conceptualisation Gruber93

33
Structure of an Ontology

• Ontologies typically have two distinct
  components
• Names for important concepts in the domain
• Elephant is a concept whose members are a kind of
  animal
• Herbivore is a concept whose members are exactly
  those animals who eat only plants or parts of
  plants
• Adult_Elephant is a concept whose members are
  exactly those elephants whose age is greater than
  20 years
• Background knowledge/constraints on the domain
• Adult_Elephants weigh at least 2,000 kg
• All Elephants are either African_Elephants or
  Indian_Elephants
• No individual can be both a Herbivore and a
  Carnivore

34
Example Ontology
35
Ontology Design and Deployment

• Given key role of ontologies in the Semantic Web,
  it will be essential to provide tools and
  services to help users
• Design and maintain high quality ontologies,
  e.g.
• Meaningful all named classes can have instances
• Correct captured intuitions of domain experts
• Minimally redundant no unintended synonyms
• Richly axiomatised (sufficiently) detailed
  descriptions
• Store (large numbers) of instances of ontology
  classes, e.g.
• Annotations from web pages
• Answer queries over ontology classes and
  instances, e.g.
• Find more general/specific classes
• Retrieve annotations/pages matching a given
  description
• Integrate and align multiple ontologies

36
Ontology Languages

• Wide variety of languages for Explicit
  Specification
• Graphical notations
• Semantic networks
• Topic Maps (see http//www.topicmaps.org/)
• UML
• RDF
• Logic based
• Description Logics (e.g., OIL, DAMLOIL, OWL)
• Rules (e.g., RuleML, LP/Prolog)
• First Order Logic (e.g., KIF)
• Conceptual graphs
• (Syntactically) higher order logics (e.g., LBase)
• Non-classical logics (e.g., Flogic, Non-Mon,
  modalities)
• Probabilistic/fuzzy
• Degree of formality varies widely
• Increased formality makes languages more amenable
  to machine processing (e.g., automated reasoning)

37
Many ontology languages use object oriented
model based on

• Objects/Instances/Individuals
• Elements of the domain of discourse
• Types/Classes/Concepts
• Sets of objects sharing certain characteristics
• Relations/Properties/Roles
• Sets of pairs (tuples) of objects
• Such languages are/can be
• Well understood
• Formally specified
• (Relatively) easy to use
• Amenable to machine processing
• Description Logics are a family of such ontology
  languages

38

• What are
• Description Logics?

39
Description Logics

• A family of logics that can be used to represent
  ontologies
• Based on
• Individuals, e.g., John
• Concepts, e.g., Person, Parent
• Roles, e.g, childOf
• Concepts can have defining characteristics
• E.g., Parent is precisely those Persons who have
  a child
• Description Logics are
• Well understood
• Formally specified with model-theoretic semantics
• (Relatively) easy to use
• Amenable to machine processing

40
Sample Description Logic Ontology Fragment

• child ? Person ? Person
• father ? inverse(child)
• spouse ? Person ? Person
• spouse is symmetric
• Student ? Person
• Parent Person ? ?1 child
• john ? Person ? nameJohn Smith
• ltjohn,sallygt ? spouse
• sally ? nameSally Brown
• ltsally,paulgt ? father
• paul ? Person ? ? 0 spouse

41
Description Logic Inferences

• sally ? Person ? ?spousePerson
• because sally is johns spouse, spouses range
  is Person, and spouse is symmetric
• paul ? Parent
• because paul is a Person, paul is sallys
  father, father is a subrole of the inverse of
  child
• paul ? ?spouse sally
• because paul has no spouse
• mary ? ( ?spouse Person ) ? ( ?spouse ? john
  )
• because if mary has a non-Person spouse, it
  cant be john,
• because john is a Person

42
Reasoning in Description Logics

• Description Logics can be quite complex
• Boolean constructs (?, ?, ?)
• (Counted) modalities (?, ?, ?, ?)
• Concept definitions ()
• Determining inference in Description Logics can
  be difficult
• Computationally intractable (EXPTIME complete or
  worse)
• Generally decidable, however
• Nevertheless, systems exist that are effective in
  practice
• Highly optimized, tuned for normal situations
• Always return the right answer
• Almost always return very quickly
• E.g., FaCT, RACER, DLP

43
Characteristics of Description Logics

• Logics
• Formal basis
• Model theory provides meaning
• Inference can be used to determine implicit
  information
• Ontology Languages
• Provide definitions for terms (concepts)
• Provide information about individuals
• Expressive
• Can say lots of things (but not everything)
• Determining inference is difficult
• Implemented
• Powerful systems exist (FaCT, DLP, RACER)

44

• (Representation) Languages
• for the Semantic Web

45
Semantic Web Languages

• Talking the Web Talk
• Use Web syntax (XML)
• Use Semantic Web identifiers
• Uniform Resource Identifier (with optional
  fragment id)
• http//www.example.com/pfps/Winisk
• Use Semantic Web syntax (RDF/XML)
• Use Semantic Web media type (applicationxml/rdf)
• E.g., OIL Ontology Inference Layer

46
Semantic Web Languages

• Walking the Web Walk
• Be compatible with RDF semantics
• Be able to handle multiple documents
• I.e., dont think of the Web as a single source
  of information
• Be able to handle inconsistency
• Worry about size
• E.g., OWL Web Ontology Language

47
Initial Semantic Web Languages

• RDF (Resource Description Framework)
• W3C recommendation (http//www.w3.org/RDF)
• RDF is a language ( XML syntax semantics)
• for representing metadata
• for describing the semantics of information in a
  machine- accessible way
• RDFS (RDF Schema) extends RDF with schema
  vocabulary
• Class, Property
• type, subClassOf, subPropertyOf
• range, domain
• RDFS is a very simple ontology language

48
The RDF Data Model

• Statements are ltsubject, predicate, objectgt
  triples
• ltIan,hasColleague,Uligt
• Can be represented as a graph

• Statements describe properties of resources
• A resource is any object that can be pointed to
  by a URI
• a document, a picture, a paragraph on the Web
• http//www.cs.man.ac.uk/index.html
• a book in the library, a real person (?)
• isbn//5031-4444-3333
• Properties themselves are also resources (URIs)

49
URIs

• URI Uniform Resource Identifier
• "The generic set of all names/addresses that are
  short strings that refer to resources"
• URLs (Uniform Resource Locators) are a particular
  type of URI, used for resources that can be
  accessed on the WWW (e.g., web pages)
• In RDF, URIs typically look like normal URLs,
  often with fragment identifiers to point at
  specific parts of a document
• http//www.somedomain.com/some/path/to/filefragme
  ntID

50
Linking Statements

• The subject of one statement can be the object of
  another
• Such collections of statements form a directed,
  labeled graph
• Note that the object of a triple can also be a
  literal (a string)

51
RDF Syntax

• RDF has an XML syntax
• Every Description element describes a resource
• Every attribute or nested element inside a
  Description is a property of that Resource
• We can refer to resources by using URIs
• ltDescription about"some.uri/person/ian_horrocks"
  gt
• lthasColleague resource"some.uri/person/uli_sa
  ttler"/gt
• lt/Descriptiongt
• ltDescription about"some.uri/person/uli_sattler"gt
  
• lthasHomePagegthttp//www.cs.mam.ac.uk/sattlerlt
  /hasHomePagegt
• lt/Descriptiongt
• ltDescription about"some.uri/person/carole_goble"
  gt
• lthasColleague resource"some.uri/person/uli_sa
  ttler"/gt
• lt/Descriptiongt

52
RDF Schema (RDFS)

• RDF gives a language for meta data annotation,
  and a way to write it down in XML, but it does
  not provide any way to structure the annotations
• RDF Schema augments RDF to allow you to define
  vocabulary terms and the relations between those
  terms
• it gives extra meaning to particular RDF
  predicates and resources
• e.g., Class, subClassOf, domain, range
• These terms are the RDF Schema building blocks
  (constructors) used to create vocabularies
• ltPerson,type,Classgt
• lthasColleague,type,Propertygt
• lthasColleague,range,Persongt
• lthasColleague,domain,Persongt
• ltProfessor,subClassOf,Persongt
• ltCarole,type,Professorgt
• ltCarole,hasColleague,Iangt

53
RDF and RDFS circa 2001

• Initial definition of RDF and RDFS was informal
• A document giving an English description of what
  everything meant
• Not adequate for representation
• Debate on exact meaning of constructs, e.g.,
  blank nodes
• Similar to problems with informal Knowledge
  Representation work
• W3C chartered the RDF Core Working Group to fix
  this (and other problems)
• Produced cleaned up syntax for RDF
• Produced formal semantics
• RDF and RDFS are now real representation
  languages
• Formal syntax, formal semantics, inference

54
Semantics and Model Theories

• Ontology/KR languages aim to model (part of)
  world
• Terms in language correspond to entities in world
• Meaning given by a Model Theory (MT)
• MT defines relationship between syntax and
  interpretations
• Can be many interpretations (models) of one piece
  of syntax
• Models are supposed to be analogue of (part of)
  world
• E.g., elements of model correspond to objects in
  world
• Formal relationship between syntax and models
• Structure of models reflect relationships
  specified in syntax
• Inference (e.g., subsumption) is defined in terms
  of MT
• E.g., T ² A ? B iff in every model of T, ext(A) ?
  ext(B)

55
RDF/RDFS Semantics

• RDF has non-standard semantics to deal with
  certain bits of RDF
• Semantics given by RDF Model Theory (MT)
• In RDF MT, an interpretation I of a vocabulary V
  consists of
• IR, a non-empty set of resources
• IS, a mapping from V into IR
• IP, a distinguished subset of IR (the properties)
• A vocabulary element v 2 V is a property iff
  IS(v) 2 IP
• IEXT, a mapping from IP into the powerset of
  IRIR
• I.e., a set of elements ltx,ygt, with x,y elements
  of IR
• IL, a mapping from typed literals into IR
• Class interpretation ICEXT simply induced by
  IEXT(IS(type))
• ICEXT(C) x ltx,Cgt 2 IEXT(IS(type))

56
RDFS Interpretations

• RDFS adds extra constraints on interpretations
• E.g., interpretations of ltC,subClassOf,Dgt
  constrained to those where ICEXT(IS(C)) µ
  ICEXT(IS(D))
• Can deal with triples such as
• ltSpecies,type,Classgt
  ltLion,type,Speciesgt
  ltLeo,type,Liongt
• ltSelfInst,type,SelfInstgt
• And even with very unusual triples such as
• lttype,subPropertyOf,subClassOfgt
• But not clear if meaning of unusual triples
  matches intuition (if there is one)

57
Problems with RDFS

• RDFS too weak to describe resources in sufficient
  detail
• No localised range and domain constraints
• Cant say that the range of hasChild is person
  when applied to persons and elephant when applied
  to elephants
• No existence/cardinality constraints
• Cant say that all instances of person have a
  mother that is also a person, or that persons
  have exactly 2 parents
• No transitive, inverse or symmetrical properties
• Cant say that isPartOf is a transitive property,
  that hasPart is the inverse of isPartOf, or that
  touches is symmetrical
• Need to extend RDFS to provide better support for
  ontologies
• Difficult to provide reasoning support
• No native reasoners for non-standard semantics
• May be possible to reason via FO axiomatisation

58
An Ontology Language for the Semantic Web

• Create a richer ontology language for the
  Semantic Web
• Desirable features identified for Web Ontology
  Language
• Extend existing Web standards
• Such as XML, RDF, RDFS
• Easy to understand and use
• Should be based on familiar KR idioms
• Formally specified
• Of adequate expressive power
• Possible to provide automated reasoning support

59
(In)famous Layer Cake
? Semanticsreasoning
?
? Relational Data
?
? Data Exchange

• Relationship between layers is not clear

60
From RDF to OWL

• Two languages were developed to satisfy above
  requirements
• OIL developed by group of (largely) European
  researchers (several from EU OntoKnowledge
  project)
• DAML-ONT developed by group of (largely) US
  researchers (in DARPA DAML programme)
• Efforts merged to produce DAMLOIL
  http//www.daml.org/language/
• Development was carried out by Joint EU/US
  Committee on Agent Markup Languages
• DAMLOIL extends (DL subset of) RDF
• DAMLOIL submitted to W3C as basis for
  standardisation
• Web-Ontology (WebOnt) Working Group formed
• WebOnt group developed OWL language based on
  DAMLOIL
• OWL language now a W3C Recommendation

61
DAMLOIL Characteristics

• Web language for Ontologies
• Uses Web syntax (XML, RDF/XML)
• Uses URI references as identifiers
• Has a notion of Web documents
• Has formal semantics and decidable inference
• Can be implemented (very close to languages
  already supported)
• Not completely integrated with Semantic Web
• Formal semantics not related to formal semantics
  of RDF
• Because RDF semantics had not yet been developed!
• Therefore work needed to be done to integrate RDF
  and ontology language semantics

62
The OWL Language (Overview)

• OWL is an ontology language, based on ideas from
  Description Logics
• Well defined semantics
• OWL extends the expressive power of RDFS
• Can talk about defined properties of classes
• OWL is not a full first-order language
• OWL DL (the Description Logic subset of OWL)
  benefits from many years of DL research
• Formal properties well understood (complexity,
  decidability)
• Known reasoning algorithms
• Implemented systems (highly optimised)

63
The OWL Language (Details)

• Three species of OWL
• OWL full is union of OWL syntax and RDF
• OWL DL restricted to well-behaved fragment (¼
  DAMLOIL)
• OWL Lite is easier to implement subset of OWL
  DL
• Semantic layering
• OWL DL ¼ OWL full within DL fragment
• DL semantics officially definitive
• OWL DL based on SHIQ Description Logic
• In fact it is equivalent to SHOIN(Dn) DL
• OWL DL Benefits from many years of DL research
• Well defined semantics
• Formal properties well understood (complexity,
  decidability)
• Known reasoning algorithms
• Implemented systems (highly optimised)

64
OWL Ontology Fragment

• OWL class definition in RDF/XML syntax (taken
  from OWL wine and food ontology)
• ltowlClass rdfID"WhiteWine"gt
• ltowlintersectionOf rdfparseType"Collection"gt
• ltowlClass rdfabout"Wine" /gt
• ltowlRestrictiongt
• ltowlonProperty rdfresource"hasColor" /gt
• ltowlhasValue rdfresource"White" /gt
• lt/owlRestrictiongt
• lt/owlintersectionOfgt
• lt/owlClassgt
• In other (more readable) words
• A WhiteWine is precisely a Wine that has color
  White.
• (From now on, I will not use this RDF/XML syntax!)

65
OWL Ontology Fragment

• wineWine ? foodPotableLiquid ? ?1
  winemadeFromGrape) ?
• (1 winehasMaker) ? ?
  winehasMaker wineWinery) ?
• (1 winehasColor) ? ...
• winemadeFromGrape ?wineWine ? wineWineGrape
• winehasColor ?wineWine ? wineWineColor
• wineWineColor wineWhite, wineRose,
  wineRed
• wineWhite ? wineRose wineWhite ? wineRed
  wineRose ? wineRed
• wineWineDescriptor wineWineTaste ?
  wineWineColor
• wineWhiteWine wineWine ? ( winehasColor
  wineWhite )
• wineRiesling wineWine ? ?1 winemadeFromGrape)
  ?
• (winemadeFromGrape
  wineRieslingGrape )
• wineRiesling ? winehasColor wineWhite
• wineCorbansDryWhiteRiesling ? wineRiesling ?
• 
  (winehasMaker wineCorbans) ? ...

66
Consequences of the Fragment

• OWL is a logic, with an entailment (consequence)
  relationship.
• For example, the ontology fragment above entails
• wineRiesling ? wineWhiteWine
• wineRiesling ? wineWhiteWine ? (1
  winehasMaker)
• wineRiesling ? (winehasColor wineRed) ?
  wineWineDescriptor
• wineCorbansDryWhiteRiesling ? wineWhiteWine
• Determining consequence in OWL is difficult
• NEXPTIME complete for the DL fragment of OWL
• Sound and complete reasoners not (quite) yet
  practical''

67
What You Can't Say in OWL

• Certain kinds of disjunction
• Either Saddam is a terrorist or Bush is a
  liar''
• But can say Bush is either a terrorist or a
  liar''
• exBush ? exTerrorist ? exLiar
• Many kinds of universal quantification
• For every student and university, there is an
  application by that student to that university''
• But can say Every student applies to at least
  five universities''
• exStudent ? (? 5 exapplication) ?
  (?exapplication exUniversity)
• Many kinds of relationships between properties
• Uncle is the composition of parent and
  brother''
• Limitations are in place to keep DL fragment of
  OWL decidable

68
What Can be Done in OWL

• Build ontologies
• What is an ontology? (See below)
• State facts (including disjunctive and vague
  facts)
• What is the difference between an ontology and a
  bunch of
  facts? (Not much!)
• OWL specifies the consequences of an ontology
• What sorts of consequences? (Whatever can be said
  in OWL.)
• OWL specifies the consequences of a bunch of
  facts
• What sorts of consequences? (Whatever can be said
  in OWL)
• Is determining consequence different from
  retrieval? (Yes, in some sense, as consequence
  provides YES/NO answers.)

69
OWL Class Constructors

• XMLS datatypes as well as classes in 8P.C and
  9P.C
• E.g., 9hasAge.nonNegativeInteger
• Arbitrarily complex nesting of constructors
• E.g., Person u 8hasChild.Doctor t 9hasChild.Doctor

70
RDF Syntax for OWL

• ltowlClassgt
• ltowlintersectionOf rdfparseType"
  collection"gt
• ltowlClass rdfabout"Person"/gt
• ltowlRestrictiongt
• ltowlonProperty rdfresource"hasChild"/gt
• ltowltoClassgt
• ltowlunionOf rdfparseType" collection"gt
• ltowlClass rdfabout"Doctor"/gt
• ltowlRestrictiongt
• ltowlonProperty rdfresource"hasChil
  d"/gt
• ltowlhasClass rdfresource"Doctor"/gt
  
• lt/owlRestrictiongt
• lt/owlunionOfgt
• lt/owltoClassgt
• lt/owlRestrictiongt
• lt/owlintersectionOfgt
• lt/owlClassgt

E.g., Person ? 8hasChild.Doctor ?
9hasChild.Doctor
71
OWL Axioms

• Axioms (mostly) reducible to inclusion (v)
• C D iff both C v D and D v C

72
XML Schema Datatypes in OWL

• OWL supports XML Schema primitive datatypes
• E.g., integer, real, string,
• Strict separation between object classes and
  datatypes
• Disjoint interpretation domain DD for datatypes
• For a datavalue d, dI µ DD
• And DD Å DI
• Disjoint object and datatype properties
• For a datatype propterty P, PI µ DI DD
• For object property S and datatype property P,
  SI Å PI
• Equivalent to the (Dn) in SHOIN(Dn)

73
Why Separate Classes and Datatypes?

• Philosophical reasons
• Datatypes structured by built-in predicates
• Not appropriate to form new datatypes using
  ontology language
• Practical reasons
• Ontology language remains simple and compact
• Semantic integrity of ontology language not
  compromised
• Implementability not compromised can use hybrid
  reasoner
• Only need sound and complete decision procedure
  for
• dI1 Å Å dIn, where d is a (possibly negated)
  datatype

74
OWL DL Semantics

• Mapping OWL to equivalent DL (SHOIN(Dn))
• Facilitates provision of reasoning services
  (using DL systems)
• Provides well defined semantics
• DL semantics defined by interpretations I (DI,
  I), where
• DI is the domain (a non-empty set)
• I is an interpretation function that maps
• Concept (class) name A ! subset AI of DI
• Role (property) name R ! binary relation RI over
  DI
• Individual name i ! iI element of DI

75
DL Semantics

• Interpretation function I extends to concept
  expressions in an obvious(ish) way, i.e.

76
DL Knowledge Bases (Ontologies)

• An OWL ontology maps to a DL Knowledge Base K
  hT , Ai
• T (Tbox) is a set of axioms of the form
• C v D (concept inclusion)
• C D (concept equivalence)
• R v S (role inclusion)
• R S (role equivalence)
• R v R (role transitivity)
• A (Abox) is a set of axioms of the form
• x 2 D (concept instantiation)
• hx,yi 2 R (role instantiation)
• Two sorts of Tbox axioms often distinguished
• Definitions
• C v D or C D where C is a concept name
• General Concept Inclusion axioms (GCIs)
• C v D where C in an arbitrary concept

77
Knowledge Base Semantics

• An interpretation I satisfies (models) an axiom A
  (I ² A)
• I ² C v D iff CI µ DI
• I ² C D iff CI DI
• I ² R v S iff RI µ SI
• I ² R S iff RI SI
• I ² R v R iff (RI) µ RI
• I ² x 2 D iff xI 2 DI
• I ² hx,yi 2 R iff (xI,yI) 2 RI
• I satisfies a Tbox T (I ² T ) iff I satisfies
  every axiom A in T
• I satisfies an Abox A (I ² A) iff I satisfies
  every axiom A in A
• I satisfies an KB K (I ² K) iff I satisfies both
  T and A

78
Inference Tasks

• Knowledge is correct (captures intuitions)
• C subsumes D w.r.t. K iff for every model I of K,
  CI µ DI
• Knowledge is minimally redundant (no unintended
  synonyms)
• C is equivallent to D w.r.t. K iff for every
  model I of K, CI DI
• Knowledge is meaningful (classes can have
  instances)
• C is satisfiable w.r.t. K iff there exists some
  model I of K s.t. CI ?
• Querying knowledge
• x is an instance of C w.r.t. K iff for every
  model I of K, xI 2 CI
• hx,yi is an instance of R w.r.t. K iff for,
  every model I of K, (xI,yI) 2 RI
• Knowledge base consistency
• A KB K is consistent iff there exists some model
  I of K
• Performing these tasks in OWL DL is difficult
  (intractable) but possible (decidable)

79
OWL as a Semantic Web Language

• Talking the Web Talk
• OWL uses URI references as identifiers
• OWL has an XML syntax
• OWL syntax is compatible with RDF syntax
• OWL uses media type application/xmlrdf

80
OWL as a Semantic Web Language

• Walking the Web Walk
• OWL is compatible with RDF semantics
• OWL has interdocument references
• Can pick and choose which documents to use
• Dont need to consider the entire Web
• Handling inconsistency is delegated to
  higher-level components
• Component that picks which documents to start with

81
OWL as a Knowledge Representation Language

• OWL is an ontology language
• Provides constructs for defining concepts and
  roles
• Provides constructs for providing information
  about individuals
• Based on Description Logics
• OWL has a formal semantics
• Determines meaning of OWL constructs
• Provides notions of inference
• OWL is can be effectively implemented
• Similar to Description Logics underlying FaCT,
  DLP, RACER

82
Conclusion

• OWL is an existence proof that the Semantic Web
  and Knowledge Representation can help each other
• Knowledge Representation provides expressive
  power, formal rigor, ways around computational
  difficulties
• The Web provides syntax, global name space,
  documents, locality

83
Making the Semantic Web better for Knowledge
Representation

• The Layer Cake vision means that problems with
  lower layers show up in all higher layers
• Same syntax together with compatible semantics
  for all languages is a particular problem
• Allow instead multiple, coexisting languages
• All talk the Web talk
• XML syntax, URI references, media types
• All also walk the Web walk
• Compatible semantics
• Interdocument references
• Use the language(s) that are appropriate for the
  task

84
The End?

• Not nearly!
• This story is not finished
• Semantic Web just starting
• OWL tools not nearly mature
• Need
• Users willing to work in an experimental area
• Developers more tools for OWL, including
  ontology-development environments and better
  interfaces