#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
Semantic Web - Ontologies
1. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Semantic Web
Unit 5: Ontologies & Linked Data
2. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 2
Semantic Web Roadmap:
Controlled growth bottom
up according to this
architecture.
Architecture was (slightly)
modified in the last years.
5.1. Why is RDF not sufficient?
3. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 3
5.1. Sharing a conceptualization
5.2. Ontologies in Computer-Science
5.3. Ontology Language
5.4. Types of Ontologies
5.6. Ontology Tools
5.7. References
5.1. Why is RDF not sufficient?
5.5. Linked Data
4. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 4
Level of knowledge representation and semantics
XML / XML Schema
objects, structure
RDF / RDF Schema
knowledge about
objects, relations
between objects
OWL
domain knowledge,
interconnections
5.1. Sharing a conceptualization
5. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 5
woman
picture
nudity
photo
artwork
porn
artwork
woman
Different people, different perceptions
Users
Author
Resource
collaborative tagging
Web 2.0 approach
authoritative metadata
Semantic Web approach
Search engine
?
Need of a shared
conceptualization
5.1. Sharing a conceptualization
6. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 6
woman
human
isa
man
isa
≠
photo
picture
is a
street
taken
1 0..*
depicted
0..*
0..*
Conceptualization
concepts
relations between
concepts
attributes
name
age
size
name
length
instances
bd. JFK
3 km
Louise Ciccone
54
173 cm
Using the same
ontology allows two
different systems to
communicate and to
reason over (meta)data
5.1. Sharing a conceptualization
7. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 7
5.2. Ontologies in Computer-Science
Ontologies in Computer-Science
An ontology has a common language (symbols, expressions ) syntax
The meaning of the symbols and expressions in an ontology is clear semantics
Symbols and expressions with similar semantics are grouped in classes conceptualization
Concepts are organized in a hierarchical way taxonomy
Implicit knowledge can be made explicit reasoning
An ontology is an explicit, formal specification of a
shared conceptualization (Thomas R. Gruber, 1993)
Conceptualization : abstract model of domain related expressions
Specification : domain related
Explicit : semantics of all expressions is clear
Formal : machine-readable
Shared : consensus (different people have different perceptions)
8. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Classes: (concepts) are abstract groups, sets, or collections of objects (individuals and classes). Here: Thing,
Human, Father, etc are classes.
Taxonomy: hierarchical representation of classes
Individuals: (instances) are the basic, "ground level" components of
an ontology. For example: SerLinck is an individual of the class Man,
formally: Man(SerLinck)
Attributes: (properties) describing objects
(individuals and classes) in the ontology. Here, the
class Human has an attribute hasName and the
individual SerLinck has the attribute value "Serge
Linckels" for the attribute hasName
Structure of ontologies
5.2. Ontologies in Computer-Science
9. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Structure of ontologies
Relationships: (associations) expressing how objects in the ontology are related to each other. Typically a
relation is an attribute whose value is another object in the ontology.
subsumption relation: (is-superclass-of) defines
which objects are members of classes. Here Man
subsumes Father.
There are two common types of relations: the vertical "subsumption"
and (normally) horizontal user defined relations.
Relations can be recursive, e.g, a human has a child
that is human.
user defined: defines any kind of relation between
objects. E.g., hasHusband is a relation from the
class Woman to the class Man.
hasChild
hasHusband
5.2. Ontologies in Computer-Science
10. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Structure of ontologies
Restrictions can be attached to relations:
quantified restrictions, e.g.,
• a woman can have 0 or 1 husband
• a human can have 0 or n children
• every mother must have at least one child
difference, e.g.,
a woman is not a man (a human can be either a
woman or a man, not both)
hasChild
hasHusband
5.2. Ontologies in Computer-Science
11. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Reasoning over ontologies
Axioms are knowledge definitions in the ontology that was explicitly defined and that have not to be proven
true
Examples:
• SerLinck is an individual of the class Father
• MagLinck is an individual of the class Woman
• MagLinck has BobLinck as child
Examples:
• Because SerLinck is an individual of the
class Man, he is human (because Human
subsumes Man)
• Because MagLinck has a child, she is an
individual of the class Mother
• The class Wife can be inductively defined as
being all the women who have at least one
husband
hasChild
hasHusband
Implicit knowledge can be made explicit by logical
induction reasoning over the ontology
5.2. Ontologies in Computer-Science
12. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 12
5.3. Ontology Language
Knowledge representation
How to represent such expressions in a computer-readable way, in order to reason over that
knowledge?
Examples of natural language:
• a woman can have 0 or 1 husband
• every mother must have at least one child
• a woman is not a man (a human can be either a woman or a man, not both)
Informal representation of knowledge:
Beside the structural dimension of an ontology, an ontology uses a common language to
formalize its specifications and conceptualizations
Ontology Language
Examples of ontology languages:
• Web Ontology Language (OWL)
• Ontology Interface Layer (OIL)
• DARPA Agent Markup Language (DAML)
• CycL
• Knowledge Interchange Format (KIF)
Most of these languages are
based on a subset of First Order
Logic (FOL)
13. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
5.3. Ontology Language
Example of knowledge representation
Informal Formal
A woman can only have male husband Woman hasHusband.Man
Every mother must have at least one human
child
Mother Woman hasChild.Human
A human can either be a woman or a man,
not both
Human Woman Man
Woman Man
Man Woman
Formal representation: computer-readable and free of ambiguities (only one interpretation
possible), e.g., code in a programming language
Informal representation: not formal, meaning something not characterized by a clear and
unambiguous interpretation, e.g., natural language
Description Logics
14. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 14
5.4. Types of Ontologies
Upper Ontologies
An upper ontology (or world ontology) is a model of the common objects that are generally
applicable across a wide range of domain ontologies
It contains a core glossary in whose terms objects in a set of domains can be described
Dublin Core metadata element set is a standard for cross-domain information resource
description. In other words, it provides a simple and standardized set of conventions for
describing things online in ways that make them easier to find.
Examples:
The General Formal Ontology (GFO) integrates processes and objects. GFO provides
a framework for building custom, domain-specific ontologies
OpenCyc includes hundreds of thousands of terms along with millions of assertions
relating the terms to each other. One stated goal is that of providing a completely free
and unrestricted semantic vocabulary for use in the Semantic Web.
Suggested Upper Merged Ontology (SUMO) was developed within the IEEE Standard
Upper Ontology Working Group. The goal is to develop a standard ontology that will
promote data interoperability, information search and retrieval, automated inferencing,
and natural language processing.
15. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 15
5.4. Types of Ontologies
Domain Ontologies
A domain ontology models a specific domain, or part of the world
It represents the particular meanings of terms as they apply to that domain
One of the most cited ontologies is the wine ontology it is about the most appropriate
combination of wine and meals
Examples:
The soccer ontology describes most concepts that are specific to soccer: players,
rules, field, supporters, actions, etc. It is used to annotate videos in order to produce
personalized summary of soccer matches
An ontology library for lung pathology is maintained by the FU-Berlin. The aim of the
project "A Semantic Web for Pathology" is to realize a semantic web based retrieval
system for the domain of lung pathology. For this purpose the pathology data is
annotated with semantic references, and the textual pathology reports are used as
descriptions of what the associated images represent
The music ontology provides main concepts and properties for describing music, i.e.
artists, albums, tracks, but also performances, arrangements, etc.
16. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 16
5.4. Types of Ontologies
Expressivity of ontologies
In General, the more specific the ontology is, the more expressive it becomes
catalog
ID
terms
glossary
thesaurus
informal
"is-a"
– +Expressivity
lightweight ontologies
controlled, unambiguous, and finite set of vocabulary in a catalog
(Lassila/McGuinness, 2001)
finite list of terms and meaning in natural language
additional semantics with relations between terms (thesaurus)
conceptualization in a hierarchy of few top-classes
17. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 17
5.4. Types of Ontologies
Expressivity of ontologies
In General, the more specific the ontology is, the more expressive it becomes
catalog
ID
terms
glossary
thesaurus
informal
"is-a"
formal
"is-a"
formal
"instance"
frames
properties
value
restrictions
disjointness
inverse
part-of
general logic
constraints
– +Expressivity
heavyweight ontologies
taxonomies with strict subclass relationships
logical induction over instance checking
classes include property information
using logical quantifiers to express restrictions
(Lassila/McGuinness, 2001)
more complex restrictions,
e.g., disjointness
complete and complex
logical expressions
17
18. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 18
5.4. Types of Ontologies
Expressivity of ontologies
In General, the more specific the ontology is, the more expressive it becomes
catalog
ID
terms
glossary
thesaurus
informal
"is-a"
formal
"is-a"
formal
"instance"
frames
properties
value
restrictions
disjointness
inverse
part-of
general logic
constraints
– +Expressivity
XML
RDF
OWL
19. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 19
Example: Thesaurus of English words
5.4. Types of Ontologies
http://www.visualthesaurus.com/
20. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 20
Example: Linnaen Taxonomy
Linnaean taxonomy is a method of classifying living
things in a taxonomy based on "is-a" relationships
By Carl Linnaeus (1707 – 1778)
5.4. Types of Ontologies
21. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 21
Example: WordNet – lexical database (thesaurus & taxonomy)
5.4. Types of Ontologies
http://wordnet.princeton.edu/
Semantically equivalent words (synsets) are interlinked by means of conceptual-semantic and
lexical relations
hyperonym: a word with a more general meaning
(e.g., animal is a hyperonym of cat),
hyponym: a word with a more specific meaning (e.g.,
cat is a hyponym of animal),
synonym: a word with identical meaning (e.g., car
and automobile are synonyms),
homonym: words with identical spelling but different
meaning (e.g., Ada is a programming language but
also a person).
22. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 22
Example: KR Ontology (Upper Ontology)
5.4. Types of Ontologies
http://www.jfsowa.com/ontology/
Top level ontology with 27
concepts and interlinked (lattice)
Describes general concepts independent
from a given context
23. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 23
Example: Cyc (Upper Ontology)
5.4. Types of Ontologies
http://www.opencyc.org/
Includes hundreds of thousands of terms along
with millions of assertions relating the terms to
each other
Complex queries can be expressed, also in natural
language
24. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 24
Example: UMBEL (Upper Ontology)
5.4. Types of Ontologies
http://www.umbel.org/
Upper Mapping and Binding Exchange Layer
Subset of OpenCyc
28000 concepts to provide commen mapping
points for relating different ontologies to one
another
Shared vocabulary for ontology mapping
Used in Linked data to to link classes of different
sub-ontologies to other datasets, e.g. 48000
mappings to DBpedia
25. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 25
Example: Wine Ontology (Domain Ontology)
5.4. Types of Ontologies
http://www.w3.org/TR/owl-guide/wine.rdf
It is about finding the most appropriate combination of wine and meals
26. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 26
"Using the Web to connect
related data that wasn't
previously linked, or using the
Web to lower the barriers to
linking data currently linked
using other methods."
Linked Open Data:
DBPedia plays a
central role as it
makes the content of
Wikipedia available in
RDF
27. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Linking open-data community project
Goal: “expose” open datasets in
RDF
Set RDF links among the
data items from different
datasets
Set up SPARQL
endpoints
Billions of triples, millions
of “links”
DBpedia is a community effort (1) to extract structured information from Wikipedia, (2) to
provide a SPARQL endpoint to the dataset, and (3) to interlink the DBpedia dataset with
other datasets on the Web
5.5. Linked Data
Inconveniences:
• incomplete
• sometimes inconsistent
28. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
5.5. Linked Data
Semantic Web ::: Serge Linckels, 2011 ::: http://www.linckels.lu/ ::: 28
Used in the IBM Watson artificial intelligence system
Knowledge base developed at the Max Planck Institute for
Computer Science in Saarbrücken
Data is automatically extracted from Wikipedia and other sources
Linked to the DBpedia ontology and uses SPARQL
29. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
SKOS - Simple Knowledge Organization System
Data model based on RDF & RDF-S for
sharing and linking knowledge
• skos:Concept defines a new concept
• skos:broader and skos:narrow
defines that a concept is more general
or more specific than another
• skos:related defines a similar
concept
• skos:exactMatch defines that two
concepts are identical (i.a.,
owl:sameAs)
5.5. Linked Data
http://www.w3.org/2004/02/skos/
30. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
DBpedia
Extracting structured data from Wikipedia
@prefix dbpedia <http://dbpedia.org/resource/>.
@prefix dbterm <http://dbpedia.org/property/>.
dbpedia:Amsterdam
dbterm:officialName "Amsterdam" ;
dbterm:longd "4" ;
dbterm:longm "53" ;
dbterm:longs "32" ;
dbterm:website <http://www.amsterdam.nl> ;
dbterm:populationUrban "1364422" ;
dbterm:areaTotalKm "219" ;
dbterm:hometown 2_Unlimited ;
dbterm:location Anne_Frank_House ;
...
New entry points
(resources)
5.5. Linked Data
31. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Automatic links among open datasets
<http://dbpedia.org/resource/Amsterdam>
owl:sameAs <http://rdf.freebase.com/ns/...> ;
owl:sameAs <http://sws.geonames.org/2759793> ;
...
<http://sws.geonames.org/2759793>
owl:sameAs <http://dbpedia.org/resource/Amsterdam>
wgs84_pos:lat “52.3666667” ;
wgs84_pos:long “4.8833333” ;
geo:inCountry <http://www.geonames.org/countries/#NL> ;
...
5.5. Linked Data
DBpedia
32. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
5.5. Linked Data
33. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Music ontology
5.5. Linked Data
34. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
New entry points
(resources)
5.5. Linked Data
Music ontology
35. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
5.5. Linked Data
Music ontology
36. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
! False positive !
(I promise)
Slideshare
5.5. Linked Data
37. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Slideshare
5.5. Linked Data
38. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Linked Data – putting it all together
5.5. Linked Data
39. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Shared
Cache
FalconS
Sindice
Marbles
Engine
Search
Engines
Linked Data on
the Web
HTTP GET
Amazon
EC2
Linked Data – putting it all together
5.5. Linked Data
40. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 40
Example: OntoEdit
5.6. Ontology Tools
http://www.ontoknowledge.org/tools/ontoedit.shtml
Supports F-Logic, RDF-Schema and OIL
Interface to the F-Logic Inference Engine
and FaCT
41. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 41
Example: Protégé
5.6. Ontology Tools
http://protege.stanford.edu/
The Protégé platform supports two main
ways of modeling ontologies via the
Protégé-Frames and Protégé-OWL
editors
Protégé ontologies can be exported into
a variety of formats including RDF(S),
OWL, and XML Schema
Java Application; multiple plug-ins
available
Interfaces to different reasoners
42. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu 42
create the classes of the taxonomy
create properties that are related to another class (range)
create properties that have literal values
create instances of classes
and properties
Example: Protégé
44. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
http://www.ted.com/talks/manuel_lima_a_visual_history_of_human_knowledge
representation of knowlege :
formaly ass tree
but now as network
45. 5. Introduction to Ontologies
Semantic Web ::: Serge Linckels ::: www.linckels.lu ::: serge@linckels.lu
Ontologies: A Silver
Bullet for Knowledge
Management and
Electronic Commerce
Dieter Fensel
5.7. References
Handbook on Ontologies
Steffen Staab, Rudi Studer
45
Linked Data - Evolving the Web into a
Global Space
Tom Heath, Christian Bizer
E-Librarian Service
User-Friendly Semantic
Search in Digital Libraries
Serge Linckels, Christoph Meinel