See how ontologies and taxonomies can play together to reach the ultimate goal, which is the cost-efficient creation and maintenance of an enterprise knowledge graph. The knowledge modelling methodology is supported by approaches taken from NLP, data science, and machine learning.
Taxonomies and Ontologies – The Yin and Yang of Knowledge Modelling
1. Andreas Blumauer
CEO, Semantic Web Company
Dr Ian Piper
UK Director
PoolParty Semantic Suite
WEBINAR
Feb 15, 2017
Taxonomies and
Ontologies
The Yin and Yang
of Knowledge
Modelling
1
2. INTRODUCTION
2
Semantic Web
Company
founder &
CEO of
Andreas
Blumauer
developer and
vendor of
2004
founded
5.7
current
Version
active at
based on
Vienna
located
part of
Taxonomies
Knowledge
Graphs
managed
with
standard for part of
is a
>200serves customers
4. AGENDA
PART 1 (Andreas Blumauer)
● Introducing the Yin and Yang of Knowledge Modelling
● Semiotic Triangle: implicit and explicit semantics
● Knowledge Acquisition Bottleneck
● Anatomy of a Knowledge Graph
PART 2 (Ian Piper)
● Modelling knowledge with taxonomies and ontologies
● Building content knowledge graphs
4
6. Yin and Yang
of Semantic
Knowledge
Modelling
4 Yin Yang
Passive/female principle in nature Active/male principle in nature
Receive information and classify Define world and reason about it
Taxonomies Ontologies/Rules
Implicit semantics Explicit semantics
Search for ‘taxonomist’ on LinkedIn →
⅔ of found persons are female
Search for ‘ontologist’ on LinkedIn →
⅔ of found persons are male
Better to let them work together!
7. Implicit
Semantics
▸ Natural languages
▸ Ambiguity versus Universality
▸ Context information and background knowledge needed
7
Susan observes Mike on a
tower with a telescope.
8. Context is King
▸ Natural languages
▸ Ambiguity versus Universality
▸ Context information and background knowledge needed
8
- Susan and Mike are persons.
- Yesterday Michael bought a Celestron.
- If one buys something, (s)he owns it and
can use it.
- Mike and Michael is the same person.
- A Celestron is a telescope.
Susan observes Mike on a
tower with a telescope.
9. Semiotic Triangle
The level of
efficiency of an
Interpretant
depends mainly on
its ability to
correctly link a
symbol with the
object it stands for.
9
Telescope
Symbol
Object
Interpretant
10. Semiotic Triangle
The level of
efficiency of an
Interpretant
depends mainly on
its ability to
correctly link a
symbol with the
object it stands for.
10
Telescope
Symbol
Object
Interpretant
http://dbpedia.org/
resource/Telescope
11. How can various
Knowledge
Modellers build
together Strong
Artificial
Intelligence?
11 Natural languages
Taxonomies
Schemas/Ontologies
Statisticalmodels
Computational Linguists
Taxonomists
DataScientists
Ontologists
12. Knowledge Acquisition
Bottleneck
Computer (networks)
need to be programmed
with sufficient amount of
knowledge before it can
begin to learn
semi-automatically
12
Knowledge Domain
Knowledge
Modellers
Knowledge Model
semantic gap
Domain
Experts
13. How does nature
go around similar
learning
bottlenecks?
13 Bla bla
bla bla.
Bla bla
bla bla
The stove is on.
The stove is hot!
Ontological model → reasoningTaxonomical model → is-a abstractions
Bla stove
bla bla.
Bla bla
bla hot
Switched on
devices are
dangerous
devices.
Switched on devices are
dangerous, only if the
operating temperature
is above 100 degrees
and the automatic
shutdown mechanism is
broken.
The stove is on.
The stove is hot!
Statistical model/cooccurences → is related
The stove is on.
The stove is hot!
Bla bla bla bla
Bla bla bla bla.
14. Co-occurence
model
14
Reference
Corpus
- Websites
- PDF, Word, …
- Abstracts from
DBpedia
- RSS Feeds
Term 8
Term 3
Term 7
Term 8
Term 6
Term 9
Term 5
Term 10
- Relevant terms and phrases
- Relevancy of terms
- co-occurence between terms and terms
Term 1
Term 4
Term 2
15. Introducing
some explicit
semantics
▸ Taxonomies
▸ SKOS taxonomies are concept and resource-based knowledge models
15
skos:
Concept
Celestron
skos:prefLabel
skos:
Concept
skos:related
Mike
skos:prefLabel
Michael
skos:altLabel
skos:
Concept
Susan
skos:prefLabel
skos:related skos:
Concept
Scheme
skos:inScheme
skos:inScheme
Person
skos:prefLabel
skos:
Concept
Tower of Babel
skos:prefLabel
skos:
Concept
skos:broader
Telescope
skos:prefLabel
skos:related
16. Corpus analysis
results in a
network of
concepts and
terms
16
I need support to
continuously extend our
taxonomy / controlled
vocabulary!
skos:
Concept
Reference
Corpus
- Websites
- PDF, Word, …
- Abstracts from
DBpedia
- RSS Feeds
skos:
Concept
skos:
Concept
Term 1
Term 3
Term 7
Term 8
Term 6
Term 4
Term 2
Term 5
- Relevant terms and phrases
- Relevancy of concepts
- co-occurence between concepts and terms
- co-occurence between terms and terms
17. PoolParty
The Combination
of Machine
Learning & Human
Intelligence
Content Manager
Integrator
Taxonomist/
Ontologist
Thesaurus
Server
Extractor
PowerTagging
uses API
is user of
is user of
is basis of
is basis of
Index
annotates
enriches
Corpus Learning/
Semantic Analysis
CMS
extends
is basis of
analyzes
uses API
17
18. Use co-occurences
between concepts
and terms to extract
‘shadow concepts’
18 This site is a
15th-century Inca
site located 2,430
metres above sea
level. It is located
in Cusco, Peru.
It is situated on a mountain ridge above
the Sacred Valley through which the
Urubamba River flows. Most
archaeologists believe that it was built as
an estate for the Inca emperor Pachacuti.
Often mistakenly referred to as the "Lost
City of the Incas", it is the most familiar
icon of Inca civilization. The Incas built
the estate around 1450, but abandoned it
a century later at the time of the
Spanish Conquest.
Inca
site
Machu
Picchu
Cusco
Inca
empire
Inca
emperor
Peru
Spanish
Conquest
Sacred
Valley
Chankas
Lost
City
Pachacuti
In addition to explicitly used concepts and terms, Machu Picchu is
extracted from the article as a Shadow Concept. As a prerequisite,
one has to provide and analyze a representative text corpus first.
Example:
20. Ontologies:
Some more
explicit
semantics
▸ Ontologies
▸ Ontologies classify things and define more specific relations and attributes
▸ Locally and globally recognised ontologies can be combined
▸ Ontologies can have various levels of expressivity (RDFS, OWL)20
schema:
Product
Telescope
schema:name
foaf:
Person
schema:owns
Mike
foaf:nick
Michael
foaf:givenName
foaf:
Person
Susan
foaf:givenName
myOnt:observes
geo:
Spatial
Thing
Tower of Babel
skos:prefLabel
schema:
Brand
schema:brand
Celestron
schema:name
myOnt:visits
21. Reasoning
21 If someone buys a Celestron,
(s)he can use it as a telescope.
buys
uses
is ais subproperty of
22. Reasoning over
SKOS taxonomies
using OWL
22
Celestron
Telescope
Optical
device
NEXSTAR SLT
Take your
explorations to
new heights with
Celestron's
NexStarSLT.
Available with a variety of optical tubes up
to 127 mm in aperture, the NexStar SLT has
something for everyone. Beginners will
appreciate the intuive SkyAlign technology,
which makes aligning your device's
computer to the night sky as easy as
centering three bright objects in the
eyepiece. The NexStar SLT is a precision
instrument that can grow with you in the
hobby of amateur astronomy for years to
come.
I’m looking for
documents about
Optical Devices
skos:broader
skos:broader is a owl:TransitiveProperty
23. Reasoning over
SKOS taxonomies
using SPARQL 1.1
property paths
More performant!
See also: SHACL
23
Celestron
Telescope
Optical
device
NEXSTAR SLT
Take your
explorations to
new heights with
Celestron's
NexStarSLT.
Available with a variety of optical tubes up
to 127 mm in aperture, the NexStar SLT has
something for everyone. Beginners will
appreciate the intuive SkyAlign technology,
which makes aligning your device's
computer to the night sky as easy as
centering three bright objects in the
eyepiece. The NexStar SLT is a precision
instrument that can grow with you in the
hobby of amateur astronomy for years to
come.
I’m looking for
documents about
Optical Devices
skos:broader
…. WHERE ?s skos:broader+ ?o …..
24. Combine
SKOS-XL with
ontologies
▸ Use custom relations
between SKOS-XL labels24
skos-xl:
Label
Switzerland@en
skos-xl:
Label
Swiss
Confederation@en
skos-xl:altLabel
my:isPredecessor
geo:
Spatial
Thing
skos-xl:prefLabel
27. Schema data
prefLabel
Venice
prefLabel
St. Mark’s Square
altLabel
Piazza
San Marco
Peggy
Guggenheim
Museum
http://schema.org/City
http://schema.org/TouristAttraction
http://schema.org/ArtGallery
Monday through
Sunday, all day
opening
Hours
image
28. CC BY-SA 3.0
Metadata
prefLabel
Venice
prefLabel
St. Mark’s Square
altLabel
Piazza
San Marco
Peggy
Guggenheim
Museum
http://schema.org/City
http://schema.org/TouristAttraction
http://schema.org/ArtGallery
Monday through
Sunday, all day
opening
Hours
29. CC BY-SA 3.0
Taxonomies and Thesauri
prefLabel
Venice
prefLabel
St. Mark’s Square
altLabel
Piazza
San Marco
Peggy
Guggenheim
Museum
http://schema.org/City
http://schema.org/TouristAttraction
http://schema.org/ArtGallery
Monday through
Sunday, all day
prefLabel
Piazza
altLabel
Town Square
broader
related
related
opening
Hours
30. CC BY-SA 3.0
Links between internal and external data
prefLabel
Venice
prefLabel
St. Mark’s Square
altLabel
Piazza
San Marco
Peggy
Guggenheim
Museum
http://schema.org/City
http://schema.org/TouristAttraction
http://schema.org/ArtGallery
Monday through
Sunday, all day
prefLabel
Piazza
altLabel
Town Square
broader
related
related
same as
opening
Hours
31. The Peggy
Guggenheim
Collection is
a modern art museum
on the Grand Canal in
the Dorsoduro
sestiere of Venice,
Italy.
same as
CC BY-SA 3.0
Mappings to data and documents
stored in other systems
prefLabel
Venice
prefLabel
St. Mark’s Square
altLabel
Piazza
San Marco
Peggy
Guggenheim
Museum
http://schema.org/City
http://schema.org/TouristAttraction
http://schema.org/ArtGallery
Monday through
Sunday, all day
prefLabel
Piazza
altLabel
Town Square
broader
related
related
opening
Hours
56. Content
knowledge
graphs:
summary
56
A content knowledge graph
approach:
● Allows separation of concerns and reduces
dependencies
● Is a major step in development of an
enterprise knowledge graph
● Provides an incremental route from current
state
● Illustrates the benefits of the Yin and Yang of
taxonomies and ontologies
57. Meet the
PoolParty Team at
some major events
in 2017
57
June 12-14, London
MarkLogic World 2017 EMEA
> More information
Sep 11-14, Amsterdam
13th Int. Conference on Semantic Systems
> More information
Nov 6-9, Washington D.C.
KM World and
Taxonomy Bootcamp
> More information
Oct 17-18, London
Taxonomy Bootcamp
> More information
Oct 21-25, Vienna
16th Int. Semantic Web Conference
> More information