Outline Annon.

SW stack IRI XML RDF RDFS OWL

Introduction to IRI,
󰁿 󰁾󰁽 XML󰂀, RDF,
󰁿 󰁾󰁽RDFS󰂀, and OWL
󰁿 󰁾󰁽 󰂀
Part1 Part2 Part3
Semantic Web (CSC751)

Ubbo Visser

Department of Computer Science

University of Miami

September 19, 2023

Outline Annon. SW stack IRI XML RDF RDFS OWL

Outline

Announcements

Semantic Web stack

Identification of resource

Essentials of the eXtensible Markup Language (XML)

Resource Description Framework (RDF)

Resource Description Framework with Schema (RDFS)

Web Ontology Language (OWL)

Outline Annon. SW stack IRI XML RDF RDFS OWL

Announcements

Assignment - Reading
(Mandatory) Appendix A, Ch. 2 - 2.5 [HKR09].
Outline Annon. SW stack IRI XML RDF RDFS OWL

Semantic Web Stack

Tim Berners-Lee version, 2006 [BL06]

Semantic Web Stack/Cake/Layer Cake

provides the architecture of the Semantic
Web.
It is a realization of hierarchy of
languages, where each layer below provides
capabilities to immediate layer above.
Each layer is associated with standards
and specifications.
The technologies inside the red boundary
are standardized and accepted to build SW
applications.
The other layers are not clearly
standardized yet. Combinations of all the
layers realizes the SW vision.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Semantic Web Stack: layers in a nutshell

Standardized Semantic Web technologies

IRI uniquely identifies resources in the

domain, and Unicode allows to manipulate
texts in different language settings.
XML creates structured data, and
QNames resolves ambiguities.
RDF creates statements on resources.
RDFS provides a lightweight ontology
language.
OWL provides an expressive ontology
language.
SPARQL queries RDF graphs.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Semantic Web Stack : layers in a nutshell

Unrealized Semantic Web technologies

RIF/SWRL allows to describe relations

that can not be described using OWL. It is
a rule language.
Cryptography verifies SW statements are
coming from trusted sources using
appropriate digital signatures.
Trust entails statements verifying that
premises are coming from trusted sources
and relying on unifying logic and proofs.
User interfaces provides a visualization
layer to humans to use SW applications.
Outline Annon. SW stack IRI XML RDF RDFS OWL
Outline Annon. SW stack IRI XML RDF RDFS OWL

Identification of resource

Reason
We need a unambiguous way to identify things and concepts, because machines need to
process and compose information automatically.
We borrow the Web resource identification idea from the Web.
Uniform Resource Identifier (URI): theoretically distinguishes resources in the Web.
Uniform Resource Locators (URL): these are Web addresses that are used to access
online documents.
Internationalized Resource Identifier (IRI): provides the way to encode Web addresses
with Unicode.
Therefore, URLs ⊆ URIs ⊆ IRIs
Content negotiation.
Outline Annon. SW stack IRI XML RDF RDFS OWL

URIs

Format
scheme:[//authority]path[?query][#fragment]
scheme: type of URI, e.g., http, ftp, irc etc.
[//authority]: domain name.
path: some relative path.
[?query]: this is optional and provides non-hierarchical information such as parameters for
a Web service.
[#fragment]: this is optional and it is commonly used for addressing parts of the
document relative to the base URI.
Not all characters are allowed in URIs.
Outline Annon. SW stack IRI XML RDF RDFS OWL
Outline Annon. SW stack IRI XML RDF RDFS OWL

Essentials of the eXtensible Markup Language (XML)

XML
XML is a markup language recommended by the World Wide Web Consortium (W3C) for
data exchange and electronic publishing.
It provides structure to unstructured text and annotated texts.
text is data, and
additional information about data is metadata (i.e., data about data).
HTML is a popular markup language to visualize Web pages. It has tags such as
<h2>Sam</h2> with predefined semantics for the visualization (e.g., Sam).
XML tags can be chosen freely and their general meaning is not predefined. Hence, its
whole purpose is to structure the documents.
Database view: XML as a data model for semi-structured data.
Every XML document is a text document with a declaration for which XML version and
the character encoding is used. e.g.,
<?xml version=“1.0” encoding=“utf-8”?>
Outline Annon. SW stack IRI XML RDF RDFS OWL

Tree structure
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML elements and attributes

XML elements:

Attribute
Root element
There is one-and-only
outermost element called root
element.
XML elements are enclosed
with matching tag-pairs.
Empty elements can
abbreviated.
Syntactically correct XML documents are
Element names are QNames.
said to be well-formed.
XML attributes:
HTML uses fixed vocabulary with fixed
Name value pairs inside of
XML elements. meaning and used for displaying
It is an alternative means to information.
sub-elements describing data. XML uses arbitrary tags and whose
meaning is not fixed.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Namespaces
Disambiguate elements or attribute names using namespaces.

Declaration: xmlns:namespace=“<URI>”
Namespace affects from the declaration and below of the sub-tree.
Multiple declarations are possible.
If we need declaration that affects the whole document, we use a mechanism so called
Document Type Definitions (DTD). We discuss this when we talk about RDF.
We are interested in XML Schema.
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML Schema
XML allows a lot of degree of freedom in encoding information.

The causes problems when exchanging XML documents among applications.

So we need an agreement about the structure of the information, including the names of
tags and attributes, and whether certain subelements are required or not.
W3C XML Schema provides the vocabulary for this task.
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML Schema
XML Schema itself is written in XML!
XML documents are valid if it corresponds to a XML Schema.
An XML Schema is a well-formed XML document that contains XML schema definitions.
It has the root element,
<xsd:schema xmlns:xsd=“[Link] and contains element
types, which can contain attribute types, which themselves refer to predefined or
user-defined datatypes.
Datatypes are, e.g., xsd:integer, xsd:string, xsd:time, xsd:date, xsd:anyURI, xsd:ID (a
specific kind of string used as an identifier of XML elements)
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML Schema
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML Schema: user defined types

Simple types: this is obtained by restricting other types. We are not allowed to embed
elements or attribute types.
Complex types: this is obtained by grouping elements and attributes.
Type inheritance: we define new complex types from existing complex types.
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML Schema: user defined types

Outline Annon. SW stack IRI XML RDF RDFS OWL

Random thoughts
“ Any damn fool could produce a better data format than XML”
- James Clark1

e.g., SOA
Service Oriented Architecture, Web Services, WSDLs, and RESTFul services.

1 [Link]
Outline Annon. SW stack IRI XML RDF RDFS OWL
Outline Annon. SW stack IRI XML RDF RDFS OWL

Resource Description Framework (RDF)

Motivation
Let say we have to encode these sentences,

TextBook(FOST ) (1)

isPublishedBy (FOST , CRC Press). (2)

How do we uniquely represent the concept TextBook, the relation isPublishedBy and the
individuals FOST and CRC Press?
If 1 and 2 are available from two different sources, how do we merge them?
RDF provides the solution to these two problems.

Resource identification
We use URIs to represent concepts, relations, and individuals. E.g.,
FOST → [Link]
isPublishedBy → [Link]
CRC Press → [Link]
Outline Annon. SW stack IRI XML RDF RDFS OWL

RDF: W3C Recommendation 20042

RDF uses a directed graph as a data model.
The implementation uses labeled Node-Edge-Node triples.
[Link]
[Link] [Link]

RDF is a data model for

describing metadata for web pages,
structured information, and
universal, machine-readable data exchange format.
The most popular serialization mechanism is XML.

2 [Link]
Outline Annon. SW stack IRI XML RDF RDFS OWL

RDF triple
[Link]
[Link] [Link]

Subject Predicate Object

URIs and blank nodes URIs URIs, blank nodes and

literals

RDF components
URIs uniquely represent resources.
Literals are for data values.
Encoded as strings.
Meaning is interpreted by the associated datatype.
Untyped literals are treated as strings.
Blank nodes for anonymously connecting sets of triples.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Graph: sets of triples

[Link]
[Link] [Link]

[Link] [Link]

Foundation of Semantic Web Technologies CRC Press

Turtle: Terse RDF Triple language

How do we serialize a RDF graph ?

URIs in angle brackets, literals enclosed in quotes, and triples end with a period. All white spaces:
blank lines, line feeds are skipped.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Turtle shortcut: prefixes

Turtle shortcut: grouping triples with same subject, and same subject and predicate
Outline Annon. SW stack IRI XML RDF RDFS OWL

W3C recommendation: XML

XML is extensively used as the message format between heterogeneous systems.
Many programming languages provide full XML parsing libraries.
The normative syntax for RDF is based on XML syntax.
[Link]
[Link] [Link]

Subject Property
Object

Subject URI Object URI

RDF language has its own namespace.

Uses tags that belong to different namespaces.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Untyped
[Link]
[Link] [Link]

[Link] [Link]

Foundation of Semantic Web Technologies CRC Press

Untyped text is taken as free text, and it is bound by XML version and character encoding.
Subject could contain multiple properties, and
Object can be used as subject for further triples.
Datatypes can contain types from XML Schema.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Datatypes
[Link]
[Link]

[Link]
"2004-02-10"^^[Link]

"RDF Primer"^^[Link]/2001/XMLSchema#string
Outline Annon. SW stack IRI XML RDF RDFS OWL

rdf:XMLLiteral for arbitrary XML fragments

Alternative representation #1
Outline Annon. SW stack IRI XML RDF RDFS OWL

XML ENTITY to prefix URLs

Base namespace
Outline Annon. SW stack IRI XML RDF RDFS OWL

Many-valued relationships: n-ary

hasIngredient(Chutney , 1lb, GreenMango) (3)

?
Outline Annon. SW stack IRI XML RDF RDFS OWL

Using blank nodes: XML

Alternate

Using blank nodes: Turtle

Alternate
Outline Annon. SW stack IRI XML RDF RDFS OWL

Open lists: containers

This has provision to add new elements.
rdf:Seq for ordered lists, rdf:Bag for unordered list, and rdf:Alt for set of alternatives.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Close lists: collections

Outline Annon. SW stack IRI XML RDF RDFS OWL
Outline Annon. SW stack IRI XML RDF RDFS OWL

Motivation
We use RDF to represent facts:

isPublishedBy (FOST , CRC Press). (4)

How do we represent:

∃hasDaughter .Daughter ⊑ Parent (5)

Women ⊑ Person (6)
hasWife ⊑ hasSpouse (7)

These are known as terminological axioms (T-Box) or schema knowledge, and RDFS
provides a weaker schema language for modeling.
Outline Annon. SW stack IRI XML RDF RDFS OWL

RDFS common facts

RDFS is a W3C recommendation.
Every RDFS document is a valid RDF document.
We use rdfs:[Link] QName.
RDFS is a knowledge representation language or ontology language.
An ontology is a description of knowledge about a domain of interest, the core of which is
a machine-processable specification with a formally defined meaning.
RDFS is a lightweight ontology language (“A little semantics goes a long way” - James
Hendler ).

Representing things
A concept (a.k.a. class) represents a set of things. We use URIs to represent classes.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Vocabulary
Class membership:
book:uri rdf:type ex:TextBook.
An URI could have multiple memberships:
book:uri rdf:type ex:TextBook
book:uri rdf:type ex:MustRead
Classes have hierarchies (a.k.a. taxonomy): each text book is a book
ex:TextBook rdfs:subClassOf ex:Book
Evey class URI is a member of:
ex:TextBook rdf:type rdfs:Class
and,
rdfs:Class rdf:type rdfs:Class
Outline Annon. SW stack IRI XML RDF RDFS OWL

Vocabulary
rdfs:Resource : class of all resources
rdf:Property : class of all properties
rdf:XMLLiteral : we know this
rdfs:Literal : class of all literal values
rdfs:Datatype : class of all datatypes
rdf:Bag, rdf:Alt, rdf:Seq, rdf:List, rdf:nil, and rdfs:Container : for containers; open and
close
rdfs:ContainerMembershipProperty : class of constrained properties
rdfs:Statement : class of reified triples
Outline Annon. SW stack IRI XML RDF RDFS OWL

Logical consequences : deduced, or inferred or implicit knowledge

x rdf : type ex : TextBook (8)

ex : TextBook rdfs : subClassOf ex : Book (9)
|=
x rdf : type ex : Book. (10)

Logical consequences : rdfs:subClassOf is transitive

ex : TextBook rdfs : subClassOf ex : Book (11)

ex : Book rdfs : subClassOf ex : PrintMedia (12)
|=
ex : TextBook rdfs : subClassOf ex : PrintMedia. (13)
Outline Annon. SW stack IRI XML RDF RDFS OWL

Logical consequences : rdfs:subClassOf is reflexive

ex : MorningStar rdfs : subClassOf ex : EveningStar (14)

ex : EveningStar rdfs : subClassOf ex : MorningStar (15)

ex : Book rdfs : subClassOf ex : Book (16)

Property hierarchies

ex:isHappilyMarriedTo rdfs:subPropertyOf ex:isMarriedTo (17)

ex : Mary ex : isHappilyMarriedTo ex : Tom (18)
|=
ex : Mary ex : isMarriedTo ex : Tom (19)
Outline Annon. SW stack IRI XML RDF RDFS OWL

Property restrictions
We use restrictions to provide a certain restriction to URIs of subject and object.
This is done via rdfs:domain and rdfs:range.

ex : isMarriedTo rdfs : domain ex : Person (20)

ex : isMarriedTo rdfs : range ex : Person (21)

Property restrictions can be applied to datatypes:

ex : hasBirthdate rdfs : range xsd : date (22)

Outline Annon. SW stack IRI XML RDF RDFS OWL

Property restriction causes problems that might be hard to debug

: authorOf rdfs : range : TextBook (23)

: authorOf rdfs : range : StoryBook (24)

This implies that the type of : authorOf both : TextBook ∧ : StoryBook

Restriction problems

: isMarriedTo rdfs : domain : Person (25)

: isMarriedTo rdfs : range : Person (26)
: UofM rdf : type : Institute (27)
: Visser : isMarriedTo : UofM (28)
|=
: UofM rdf : type : Person (29)
Outline Annon. SW stack IRI XML RDF RDFS OWL

Reification
We want to say “The detective supposes that the butler killed the gardener”.
These are unsatisfactory:

: detective : supposes “Thebutlerkilledthegardner ′′

: detective : supposes : TheButlerKilledTheGardner

What we would like:

: butler : killed : gardener . (30)
We use reification:

: detective : supposes : id
: id rdf : subject : butler
: id rdf : predicate : hasKilled
: id rdf : object : gardener .
Outline Annon. SW stack IRI XML RDF RDFS OWL

Example
Outline Annon. SW stack IRI XML RDF RDFS OWL

RDFS semantics is weaker

:Mary rdf:type :Person ⇝ Person(Mary)
:Mother rdf:subClassOf :Woman ⇝ Mother ⊑ Woman
:John :hasWife :Mary ⇝ hasWife(John, Mary)
:hasWife rdfs:subPropertyOf :hasSpouse ⇝ hasWife ⊑ hasSpouse
:hasWife rdfs:range :Woman ⇝ ⊤ ⊑ ∀[Link]
:hasWife rdfs:domain :Man ⇝ ∃hasWife.⊤ ⊑ Man

Multiple views
Look at this statement Truck ⊑ MotorVehicle. When this statement travels up the Semantic
Stack, it will be subjected to three views:
XML structure
RDF graph (triple)
RDF Schema (semantic)
Outline Annon. SW stack IRI XML RDF RDFS OWL
Outline Annon. SW stack IRI XML RDF RDFS OWL

Assignment
Assignment #2 requires a substantial amount of reading, and modeling a simple ontology.
Please start this assignment early!

Reading
(Must read) Ch. 4 [HKR09].

Protégé
We use the Protégé ontology editor and knowledge acquisition system to demonstrate
important aspects of ontology modeling.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Motivation
How do we represent these sentences:
Every project has at least one participant.
Projects are always internal or external projects.
The superior of my superior is also my superior.
All examiners of an exam must be professors.
Human ⊑ ∃[Link]
Orphan ⊑ Human ⊓ ∀hasParent.¬Alive
Orphan(HarryPotter )
hasParent(HarryPotter , JamesPotter )
∀x, y (∃(hasParent(x, z) ∧ hasBrother (z, y )) ⇒ hasUncle(x, y ))
HappyFather ≡ ≥ [Link]
Car ⊑ =4hasTyre.⊤
PersonCommittingSuicide ≡ ∃[Link]
¬hasColleague(UbboVisser , NadalRafael)
We use OWL 2 Web Ontology Language.
OWL 2 is a W3C recommendation for modeling ontologies.
OWL Lite ⊆ OWL DL ⊆ OWL Full.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
worst-case computational complexity: NExpTime
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
worst-case computational complexity: NExpTime
OWL Lite
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
worst-case computational complexity: NExpTime
OWL Lite
contained in OWL Full and OWL DL
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
worst-case computational complexity: NExpTime
OWL Lite
contained in OWL Full and OWL DL
decidable
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
worst-case computational complexity: NExpTime
OWL Lite
contained in OWL Full and OWL DL
decidable
less expressive
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL sublanguages
OWL Full
contains OWL DL and OWL Lite
only sublanguage containing all of RDFS
very expressive
semantically difficult to understand and to work with
undecidable
support by hardly any software tools
OWL DL
contains OWL Lite and is contained in OWL Full
decidable
fully supported by most software tools
worst-case computational complexity: NExpTime
OWL Lite
contained in OWL Full and OWL DL
decidable
less expressive
worst-case computational complexity: ExpTime
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Abstract roles (a.k.a. object properties): hasAffiliation.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Abstract roles (a.k.a. object properties): hasAffiliation.
Concrete roles (a.k.a. datatype properties): firstName.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Abstract roles (a.k.a. object properties): hasAffiliation.
Concrete roles (a.k.a. datatype properties): firstName.
Domain and ranges. Use these as the last resort.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Abstract roles (a.k.a. object properties): hasAffiliation.
Concrete roles (a.k.a. datatype properties): firstName.
Domain and ranges. Use these as the last resort.
Simple class relations.
Professor ⊑ FacultyMember .
Book ⊑ Publication.
Professor ⊑ ¬Publication ≡ Professor ⊓ Publication ⊑ ⊥.
Man ⊑ Person.
Person ≡ Human.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Abstract roles (a.k.a. object properties): hasAffiliation.
Concrete roles (a.k.a. datatype properties): firstName.
Domain and ranges. Use these as the last resort.
Simple class relations.
Professor ⊑ FacultyMember .
Book ⊑ Publication.
Professor ⊑ ¬Publication ≡ Professor ⊓ Publication ⊑ ⊥.
Man ⊑ Person.
Person ≡ Human.
There is no Unique Name Assumption (UNA): owl:sameAs.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

The header of an OWL ontology.
Classes, roles, and individuals.
owl:Thing (⊤), owl:Nothing (⊥), owl:topProperty (U), and owl:bottomProperty
Classes (a.k.a. concepts): Professor
Individuals: Professor (UbboVisser )
Abstract roles (a.k.a. object properties): hasAffiliation.
Concrete roles (a.k.a. datatype properties): firstName.
Domain and ranges. Use these as the last resort.
Simple class relations.
Professor ⊑ FacultyMember .
Book ⊑ Publication.
Professor ⊑ ¬Publication ≡ Professor ⊓ Publication ⊑ ⊥.
Man ⊑ Person.
Person ≡ Human.
There is no Unique Name Assumption (UNA): owl:sameAs.
Close Classes: owl:oneOf.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Conjunction of classes: StaffOfCS ⊑ Staff ⊓ MemberOfCS.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Conjunction of classes: StaffOfCS ⊑ Staff ⊓ MemberOfCS.
Mother ≡ Woman ⊓ Parent,
∀x(Mother (X ) ⇔ Woman(x) ∧ Parent(x))

: Mother owl : equivalentClass : x.

:x rdf : type owl : Class.
:x owl : intersectionOf (: Woman : Parent).
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Conjunction of classes: StaffOfCS ⊑ Staff ⊓ MemberOfCS.
Mother ≡ Woman ⊓ Parent,
∀x(Mother (X ) ⇔ Woman(x) ∧ Parent(x))

: Mother owl : equivalentClass : x.

:x rdf : type owl : Class.
:x owl : intersectionOf (: Woman : Parent).

Disjunction of classes: Professor ⊑ ActivelyTeaching ⊔ Retired.

Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Conjunction of classes: StaffOfCS ⊑ Staff ⊓ MemberOfCS.
Mother ≡ Woman ⊓ Parent,
∀x(Mother (X ) ⇔ Woman(x) ∧ Parent(x))

: Mother owl : equivalentClass : x.

:x rdf : type owl : Class.
:x owl : intersectionOf (: Woman : Parent).

Disjunction of classes: Professor ⊑ ActivelyTeaching ⊔ Retired.

Parent ≡ Mother ⊔ Father ,
∀x(Parent(X ) ⇔ Mother (x) ∨ Father (x))

: Parent owl : equivalentClass : x.

:x rdf : type owl : Class.
:x owl : unionOf (: Mother : Father ).
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Simple class relations.
Negation: ChildlessPerson ≡ Person ⊓ ¬Parent,
∀x(ChildlessPerson(X ) ⇔ Person(x) ∧ ¬Parent(x))

: ChildlessPerson owl : equivalentClass : x.

:x rdf : type owl : Class.
:x owl : intersectionOf (: Person : y ).
:y owl : complementOf : Parent.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role restrictions.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role restrictions.
All examiners of an exam must be professors,
Exam ⊑ ∀hasExaminer .Professor
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role restrictions.
All examiners of an exam must be professors,
Exam ⊑ ∀hasExaminer .Professor
Any exam must have at least one examiner.
Exam ⊑ ∃hasExaminer .Professor
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role restrictions.
All examiners of an exam must be professors,
Exam ⊑ ∀hasExaminer .Professor
Any exam must have at least one examiner.
Exam ⊑ ∃hasExaminer .Professor
Universal quantification: only to be used with a role - a.k.a. property restrictions.
Person ⊓ Happy ≡ ∀[Link]
∀x(Person(x) ∧ Happy (x) ⇔ ∀y (hasChild(x, y ) ⇒ Happy (y )))

:x rdf : type owl : Class.

:x owl : intersectionOf (: Person : Happy ).
:x owl : equivalentClass : y.
:y rdf : type owl : Restriction.
:y owl : onProperty : hasChild.
:y owl : allValuesFrom : Parent.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Existential quantification: only to be used with a role - a.k.a. property restrictions
Parent ≡ ∃[Link]
∀x(Parent(x) ⇔ ∃y (hasChild(x, y ) ∧ Person(y )))

: Parent owl : equivalentClass : x.

:x rdf : type owl : Restriction.
:x owl : onProperty : hasChild.
:x owl : someValuesFrom : Person.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Existential quantification: only to be used with a role - a.k.a. property restrictions
Parent ≡ ∃[Link]
∀x(Parent(x) ⇔ ∃y (hasChild(x, y ) ∧ Person(y )))

: Parent owl : equivalentClass : x.

:x rdf : type owl : Restriction.
:x owl : onProperty : hasChild.
:x owl : someValuesFrom : Person.

Cardinality restrictions: at most, at least and exactly. Lets understand these constructs
using [Link]
Exam ⊑≤ 2hasExaminer .⊤
Exam ⊑≥ 3hasTopics.⊤
Exam ⊑ =3hasTopics.⊤
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
hasExaminer ⊑ hasParticipant
hasParticipant ≡ hasAttendee
hasAttendee − ≡ participatesIn
hasExaminer − ≡ examinerOf
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
hasExaminer ⊑ hasParticipant
hasParticipant ≡ hasAttendee
hasAttendee − ≡ participatesIn
hasExaminer − ≡ examinerOf
Properties can have following restrictions:
Role char. DL e.g., General presentation
Transitive Tra(R) hasAncestor R(a, b) and R(b, c) ⇒
R(a, c)
Symmetric Sym(R) hasSpouse R(a, b) ⇒ R(b, a)
Asymmetric Asy(R) hasChild R(a, b) ⇒ not R(b, a)
Reflexive Ref(R) hasRelative R(a, a) for all a
Irreflexive Irr(R) parentOf not R(a, a) for any a
Functional Fnc(R) hasHusband R(a, b) and R(a, c) ⇒
b=c
InverseFunctional Ifn(R) hasHusband R(a, b) and R(c, b) ⇒
a=c
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
Sym(hasColleague)
Tra(hasColleague)
Fun(hasTeamLeader )
Ifn(isTeamLeaderFor )
hasColleague(UbboVisser , AndreasSeekircher )
hasColleague(AndreasSeekircher , JustinStoecker )
hasColleague(JustinStoecker , SamindaAbeyruwan)
isTeamLeaderFor (UbboVisser , RoboCanes)
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
Sym(hasColleague)
Tra(hasColleague)
Fun(hasTeamLeader )
Ifn(isTeamLeaderFor )
hasColleague(UbboVisser , AndreasSeekircher )
hasColleague(AndreasSeekircher , JustinStoecker )
hasColleague(JustinStoecker , SamindaAbeyruwan)
isTeamLeaderFor (UbboVisser , RoboCanes)
Self: PersonCommittingSuicide ≡ ∃[Link]
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
Sym(hasColleague)
Tra(hasColleague)
Fun(hasTeamLeader )
Ifn(isTeamLeaderFor )
hasColleague(UbboVisser , AndreasSeekircher )
hasColleague(AndreasSeekircher , JustinStoecker )
hasColleague(JustinStoecker , SamindaAbeyruwan)
isTeamLeaderFor (UbboVisser , RoboCanes)
Self: PersonCommittingSuicide ≡ ∃[Link]
Disjoint properties, Dis(S,R) : Dis(hasParent, hasChild)
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
Sym(hasColleague)
Tra(hasColleague)
Fun(hasTeamLeader )
Ifn(isTeamLeaderFor )
hasColleague(UbboVisser , AndreasSeekircher )
hasColleague(AndreasSeekircher , JustinStoecker )
hasColleague(JustinStoecker , SamindaAbeyruwan)
isTeamLeaderFor (UbboVisser , RoboCanes)
Self: PersonCommittingSuicide ≡ ∃[Link]
Disjoint properties, Dis(S,R) : Dis(hasParent, hasChild)
Negated role assignment: ¬hasColleague(UbboVisser , NadalRafael)
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
Sym(hasColleague)
Tra(hasColleague)
Fun(hasTeamLeader )
Ifn(isTeamLeaderFor )
hasColleague(UbboVisser , AndreasSeekircher )
hasColleague(AndreasSeekircher , JustinStoecker )
hasColleague(JustinStoecker , SamindaAbeyruwan)
isTeamLeaderFor (UbboVisser , RoboCanes)
Self: PersonCommittingSuicide ≡ ∃[Link]
Disjoint properties, Dis(S,R) : Dis(hasParent, hasChild)
Negated role assignment: ¬hasColleague(UbboVisser , NadalRafael)
Role chains: hasParent ◦ hasBrother ⊑ hasUncle
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL syntax and intuitive semantics

Role relationships:
Sym(hasColleague)
Tra(hasColleague)
Fun(hasTeamLeader )
Ifn(isTeamLeaderFor )
hasColleague(UbboVisser , AndreasSeekircher )
hasColleague(AndreasSeekircher , JustinStoecker )
hasColleague(JustinStoecker , SamindaAbeyruwan)
isTeamLeaderFor (UbboVisser , RoboCanes)
Self: PersonCommittingSuicide ≡ ∃[Link]
Disjoint properties, Dis(S,R) : Dis(hasParent, hasChild)
Negated role assignment: ¬hasColleague(UbboVisser , NadalRafael)
Role chains: hasParent ◦ hasBrother ⊑ hasUncle
Datatypes (D) :
hasAge(Sam, “40”ˆˆxsd:integer)
¬hasHeight(Sam, “6.0”ˆˆxsd:float)
Teenager ≡ Person ⊓ ∃hasAge.(xsd : integer ≥ 12 and ≤ 19)
Outline Annon. SW stack IRI XML RDF RDFS OWL

Type separation, and punning, and declarations

In OWL 2 DL, a class name may also occur as a abstract role name. But, they are
treated as distinct. This is called punning.
When a class name is used as a abstract role name, they are identified by the same URI. It
is the same resource in the sense of RDF.
In OWL 2 DL, they are considered as semantically distinct, i.e., two different
interpretation of the same resource.
e.g.,
Professor (UbboVisser )
Professor (UbboVisser , UniversityOfMiami)
owl:hasKey: Given a class C , a set of abstract or concrete roles r1 , . . . , rn is said to be a
key for class C , if no two named instances of C coincide on all values of all the roles. This
relates to inverse functionality, but inverse functionality only implied the existence.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL Species
OWL Full:
Unrestricted OWL 2 DL plus all of RDF(S).
There is no reasoner that supports the semantics of OWL Full.
Type separation is not enforced. i.e., OWL Full individuals, classes, and roles can be mixed
freely. e.g., individual in one statement becomes a role in next statement.
OWL DL:
Description logic version of OWL.
Model-theoretic semantics of SROIQ(D) is used, called OWL 2 Direct Semantics.
Reasoner support exists.
OWL Lite:
OWL Lite is essentially difficult to deal with as OWL DL. Therefore, this has minor role in
practice.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL 2 Profiles
There are sublanguages of OWL 2, which have polynomial inference algorithms.
OWL 2 EL (OWL 2 EL++):
Polynomial time algorithms exist for satisfiability checking, classification, and instance
checking.
e.g., SNOMED CT
Allowed : ⊓ ∃ ⊤ ⊥ ⊑ ⊓ ∃ ⊤ ⊥, closed classes must have only one member, and
property chain axioms and range restrictions under certain conditions.
Disallowed : ¬ ⊔, arbitrary universal quantification, and role inverses.
e.g.,
Human ⊑ ∃[Link],
∃married.⊤ ⊓ CatholicPriest ⊑ ⊥,
hasParent ◦ hasParent ⊑ hasGrandparent.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL 2 Profiles
OWL 2 QL (DL LiteR ):
Answer to: what fraction of OWL 2 DL can be captured by rational database systems?
Query answering in LOGSPACE w.r.t. data via translation into SQL.
Allowed:
Domain, range, and subproperties.
Subclass statements with:
Left hand side: class name or expression of type ∃R.⊤
Right hand side: intersection of class names, expression of types ∃R.C , and negation of left
hand expressions.
No closed classes.
e.g., ∃married.⊤ ⊑ ¬Free ⊓ ∃[Link]
OWL RL (DLP) :
Answer to: what fraction of OWL 2 DL can be expressed naively by rules?
Read section [Link] of [HKR09].
Outline Annon. SW stack IRI XML RDF RDFS OWL

Ontology Vs. Database [Hor10, BHS03]

Ontology:
Open World Assumption (OWA): missing information treated as unknown.
No Unique Name Assumption (NUNA): individual may have multiple synonyms.
Ontologies provide entailments.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Ontology Vs. Database [Hor10, BHS03]

Ontology:
Open World Assumption (OWA): missing information treated as unknown.
No Unique Name Assumption (NUNA): individual may have multiple synonyms.
Ontologies provide entailments.
Database:
Close World Assumption (CWA): missing information is false.
Unique Name Assumption (UNA): each individual is uniquely identifiable.
Database schema provides structure on data.
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
T-Box:

HogwartsStudent ≡ Student ⊓ ∃[Link]

HogwartsStudent ⊑ ∀hasPet.(Owl ⊔ Cat ⊔ Toad)
hasPet ≡ isPetOf −
∃hasPet.⊤ ⊑ Human
Phoenix ⊑ ∀isPetOf .Wizard
Muggle ⊓ Wizard ⊑ ⊥
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
T-Box:

HogwartsStudent ≡ Student ⊓ ∃[Link]

HogwartsStudent ⊑ ∀hasPet.(Owl ⊔ Cat ⊔ Toad)
hasPet ≡ isPetOf −
∃hasPet.⊤ ⊑ Human
Phoenix ⊑ ∀isPetOf .Wizard
Muggle ⊓ Wizard ⊑ ⊥

A-Box:

Wizard(HarryPotter )
Wizard(DracoMalfoy )
hasFriend(HarryPotter , RonWeasley )
hasFriend(HarryPotter , HermioneGranger )
hasPet(HarryPotter , Hedwig )
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Is DracoMalfoy a friend of HarryPotter ?
Ontology: Don’t know (OWA), Database: No!
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Is DracoMalfoy a friend of HarryPotter ?
Ontology: Don’t know (OWA), Database: No!
How many friends does HarryPotter have?
Ontology: At least 1 (NUNA), Database: 2!
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Is DracoMalfoy a friend of HarryPotter ?
Ontology: Don’t know (OWA), Database: No!
How many friends does HarryPotter have?
Ontology: At least 1 (NUNA), Database: 2!
A-Box: Dis(RonWeasley , HermioneGranger )
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Is DracoMalfoy a friend of HarryPotter ?
Ontology: Don’t know (OWA), Database: No!
How many friends does HarryPotter have?
Ontology: At least 1 (NUNA), Database: 2!
A-Box: Dis(RonWeasley , HermioneGranger )
How many friends does HarryPotter have?
Ontology: at least 2, Database: 2!
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Is DracoMalfoy a friend of HarryPotter ?
Ontology: Don’t know (OWA), Database: No!
How many friends does HarryPotter have?
Ontology: At least 1 (NUNA), Database: 2!
A-Box: Dis(RonWeasley , HermioneGranger )
How many friends does HarryPotter have?
Ontology: at least 2, Database: 2!
T-Box :
HarryPottersFriends ≡ ∀hasFriend.{RonWeasley ⊔ HermioneGranger }
Wizard ⊓ HarryPottersFriends(HarryPotter )
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
Is DracoMalfoy a friend of HarryPotter ?
Ontology: Don’t know (OWA), Database: No!
How many friends does HarryPotter have?
Ontology: At least 1 (NUNA), Database: 2!
A-Box: Dis(RonWeasley , HermioneGranger )
How many friends does HarryPotter have?
Ontology: at least 2, Database: 2!
T-Box :
HarryPottersFriends ≡ ∀hasFriend.{RonWeasley ⊔ HermioneGranger }
Wizard ⊓ HarryPottersFriends(HarryPotter )
How many friends does HarryPotter have?
Ontology: 2!, Database: 2!
Outline Annon. SW stack IRI XML RDF RDFS OWL

E.g.,
If the domain or range axioms are used correctly, they provide powerful inferences:
T-Box says: ∃hasPet.⊤ ⊑ Human, Phoenix ⊑ ∀isPetOf .Wizard
A-Box:
Wizard(Dumbledore)
Phoenix(Fawkes)
isPetOf (Fawkes, Dumbledore)
Ontology infers that Human ⊓ Wizard(Dumbledore)
Database rejects, because domain of hasPet is Human, and Dumbledore is not Human
(CWA).
Ontologies use theorem proving to answer questions.
It involves both T-Box and A-Box, and has higher worst case complexities.
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL 2 features

Feature DL FOL Vocabulary

Top/bottom class ⊤/⊥ Axiomatize owl:Thing/ owl:Nothing
Class intersection ⊓ ∧ owl:intersectionOf
Class union ⊔ ∨ owl:unionOf
Class complement ¬ ¬ owl:complementOf
Enumerated {a} ≈ owl:oneOf
classes
Property restric- DL FOL Vocabulary
tion (owl:onProperty)
Existential ∃R.C ∃x . . . owl:someValuesFrom
Universal ∀R.C ∀x . . . owl:allValuesFrom
Min. cardinality ≥n R.C ∃x1 , . . . , xn owl:minQualifiedCardinality/
owl:onClass
Max. cardinality ≤n R.C ∀x1 , . . . , xn owl:maxQualifiedCardinality/
(. . . ⇒ owl:onClass
. . .)
Local reflexivity ∃[Link] R(x, x) owl:hasSelf
Outline Annon. SW stack IRI XML RDF RDFS OWL

OWL 2 features

Feature DL FOL Vocabulary

Property chain ◦ Axiomatize owl:propertyChainAxiom
Inverse R− Axiomatize owl:inverseOf
Key - Axiomatize owl:hasKey
Property disjoint- Dis(R, S) Axiomatize owl:propertyDisjointWith
ness
Property charac- DL FOL Vocabulary
teristics
Symmetric Sym(R) Axiomatize owl:SymmetricProperty
Asymmetric Asy (R) Axiomatize owl:AsymmetricProperty
Reflexive Ref (R) Axiomatize owl:ReflexiveProperty
Irreflexive Irr (R) Axiomatize owl:IrreflexiveProperty
Transitive Tra(R) Axiomatize owl:TransitiveProperty
Functional Fun(R) Axiomatize owl:FunctionalProperty
Inverse functional Ifn(R) Axiomatize owl:InverseFunctionalProperty

Subclass C ⊑D ∀x.C (x) ⇒ D(x) rdfs:subClassOf

Subproperty R⊑S ∀x, y .R(x, y ) ⇒ rdfs:subPropertyOf
S(x, y )
Outline Annon. SW stack IRI XML RDF RDFS OWL

Model these axioms in an ontology

{John} ⊓ {Bill} ⊑ ⊥
{John} ≡ {Jim}
¬hasWife(Bill, Mary )
Woman ⊓ Man ⊑ ⊥
Parent ≡ Mother ⊔ Father
ChildlessPerson ≡ Person ⊓ ¬Parent
Person ⊓ ¬Parent(Jack)
Parent ≡ ∃[Link]
Orphan ≡ ∀hasChild − .Dead
JohnsChildren ≡ ∃hasParent.{John}
NarcissisticPerson ≡ ∃[Link]
MyBirthdayGuests ≡ {Bill, John, Mary }
hasParent ◦ hasParent ⊑ hasGrandparent
Outline Annon. SW stack IRI XML RDF RDFS OWL

Acknowledgement

Acknowledgement
The majority of the slides for this course have been prepared by Saminda Abeyruwan.
Outline Annon. SW stack IRI XML RDF RDFS OWL

Franz Baader, Ian Horrocks, and Ulrike Sattler.

Description logics as ontology languages for the semantic web.
In Festschrift in honor of Jörg Siekmann, Lecture Notes in Artificial Intelligence, pages 228–248. Springer-Verlag, 2003.

Tim Berners-Lee.
Artificial Intelligence and the Semantic Web.
[Link] aaai- tbl/, 2006.

Pascal Hitzler, Markus Krötzsch, and Sebastian Rudolph.

Foundations of Semantic Web Technologies.
Chapman & Hall/CRC, 2009.

Ian Horrocks.
Description Logic: A formal foundation for languages and tools. Tutorial at the Semantic Technology Conference (SemTech). San Francisco, California, USA.
[Link] 2010.