آنتولوژی در پروتوغه

1. In The Name Of Allah

2. Instructor: Dr.Ghasem-Aghaie
Presentation: Mohammad Sadegh Salehi
Ontology Development in
Protégé

3. Protégé – Winter 2014
 Session 1: Ontology And OWL Basics,
 Syntatic Web,
 Problem of Syntatic Web,
 Semantic Web,
 Ontology and Pizza Example,
 OWL,
 Session 2: Protégé & Protégé-OWL,
 What is Protégé?
 Protégé and OWL,
 Pizza in Protégé ,
Agenda

4. Ontology and OWL

5. Protégé – Winter 2014
Hard Work using the Syntactic Web…
Find images of Tim Bray and Alan Rector…
Tim Bray
Sun
Microsystems
Rev. Alan M.
Gates,
Associate
Rector of the
Church of the
Holy Spirit,
Lake Forest,
Illinois
Prof Alan
Rector,
University of
Manchester,
Medical
Informatics
Group

6. Protégé – Winter 2014
Impossible (?) using the Syntactic Web…
Complex queries involving background knowledge
 Find information about “animals that use sonar but are not either bats
or dolphins”, e.g., Barn owl
Locating information in data repositories
 Travel enquiries
 Prices of goods and services
Finding and using “web services”
 Book me a holiday next weekend somewhere warm, not too far
away, and where they speak French or English

7. Protégé – Winter 2014
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…

8. Protégé – Winter 2014
What is the Problem?
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 7 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 …
What information can we see…

9. Protégé – Winter 2014
What is the Problem?

      
  
  
  
     
   
    

      

     

       …
What information can a machine see…

10. Protégé – Winter 2014
What is the Problem?
Solution: XML markup with “meaningful” tags?
<name>    
</name>
<location>    
</location>
<date>  </date>
<slogan>     </slogan>
<participants>   
      </participants>
<introduction>       
 
 … </introduction>
<speaker> </speaker>
<bio>       </bio>…

11. Protégé – Winter 2014
What is the Problem?
But What About… Machine sees…
<>
     </>
<>  
  </>
<>  </>
<>     </>
<>   
      
       
</>
<> 
       
  …
 </>
<> </>
<>        </>
<> </>
<>        </>

12. Protégé – Winter 2014
Need to Add “Semantics”
Use 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 formall specified
 Can also specify relationships between terms in

13. Protégé – Winter 2014
Our Domain
Pizzas have been used in Manchester
tutorials for years,
Tutorial developed by BioHealth Informatics
Group in Manchester,

14. Protégé – Winter 2014
Pizzas…
are fun
are internationally known
are highly compositional
are limited in scope
are fairly uncontroversial

15. Protégé – Winter 2014
Pizzas Ontology …
Pizza
Pizza Topping Pizza_base
Vegetarian
Pizza Vegetable
Topping
Margareta
Pizza
Tomato
Topping
Cheese
Topping
Spicy Beef
Pizza
Mozzarella
Topping
Deep
Dish base
Regular
base

16. Protégé – Winter 2014
 Is the Web Ontology Language…
 is a standard – Web Ontology Language
 comes in 3 flavours (lite, DL and full)
 we are using OWL DL (Description Logic)
 DL = decidable fragment of First Order Logic (FOL)
 Is used for capturing knowledge in a machine
understandable way
Review of OWL

17. Protégé – Winter 2014
OWL helps us …
 Describe something, rather than just name it.
 Class (BlueThing) does not mean anything
 Class (BlueThing complete
owl:Thing
restriction (hasColour someValuesFrom (Blue)))

18. Protégé – Winter 2014
OWL is a syntax independent language that has
several common representations
 – Abstract Syntax
 – N3
 – RDF/XML
OWL Syntax

19. Protégé – Winter 2014
One of the clearer human-readable syntaxes
OWL Syntax: Abstract Syntax
Class(SpicyPizza complete
annotation(rdfs:label "PizzaTemperada"@pt)
annotation(rdfs:comment "Any pizza that has a spicy topping
is a SpicyPizza"@en)
Pizza
restriction(hasTopping someValuesFrom(SpicyTopping))
)

20. Protégé – Winter 2014
OWL Constructs
Person Country
= class (concept)
Animal
= individual (instance)
Belgium
Paraguay
China
Latvia
Elvis
Hai
Holger
Kylie
S.Claus
Rudolph
Flipper
= property (relationship)

21. Protégé & Protégé-OWL

22. Protégé – Winter 2014
 Developed by Stanford Medical Informatics,
 Has a large user community (approx 300k),
 Free, open source ontology editor and knowledge-base
framework.
What is Protégé?

23. Protégé – Winter 2014
 Based on Java, is extensible,
 Provides a plug-and-play environment
 Supported by a strong community of developers and
academic, government and corporate users
What is Protégé?(Cont.)

24. Protégé – Winter 2014
Keep in mind that Protégé is...
 ...not a programming language.
 ...only a tool. It will not prevent you from making mistakes.
 Notice that we will be referring to Protégé 4.3 in this and
the following hands-on exercise.
What is Protégé?(Cont.)

25. Protégé – Winter 2014
 Create a new ontology from scratch
 Download and extend an existing ontology
 Export ontologies in a variety of formats
 OWL
 RDF
 XML Schema
What can you do with Protégé?

26. Protégé – Winter 2014
 Homepage
 http://protege.stanford.edu/
 Download
 http://protege.stanford.edu/download/protege/4.3/installanywh
ere/
 Documentation
 http://protegewiki.stanford.edu/index.php/Protege4UserDocs
 Plugins
 http://code.google.com/p/co-ode-owl-
plugins/downloads/list?can=3
Get Protégé

27. Protégé – Winter 2014
Protégé Overview

28. Protégé – Winter 2014
Classes in OWL
Classes: ‟OWL classes are interpreted as sets.”
Primitive Classes: ‟Classes that only have necessary
conditions”; Example: Animal, Cat…
Defined Classes: ‟A class that has at least one necessary
and sufficient condition”; Example: All things that are biped
and lack feathers.

29. Protégé – Winter 2014
Classes in Protégé
 All classes are subclasses of Thing
 Classes overlap by default!
 Can use tools menu to create class hierarchies

30. Protégé – Winter 2014
Properties in OWL
 Object Properties
 Relationships between two individuals
 Correspond to relationships in UML
 Examples: is_part_of, is_kissing
 Datatype Properties
 Relationships between an individual and data values
 Correspond to attributes in UML
 Examples: has_DateValue, has StringValue

31. Protégé – Winter 2014
Properties in Protégé
Object or data type
Properties can have
Subproperties
Inverse properties

32. Protégé – Winter 2014
Protégé Tabs
Protégé tabs

33. Protégé – Winter 2014
Protégé OWL plugin: Tabs
Used in this tutorial
Changing the GUI
Populating the
model
Top-level
functionality
Extensions
(visualisation)

34. Protégé – Winter 2014
Classes Tab

35. Protégé – Winter 2014
ClassesTab: Asserted Class Hierarchy
Subsumption hierarchy (superclass/subclass)
Structure as asserted by the ontology engineer
Create and Delete classes (actually subclasses!!)
Everything is a subclass of owl:Thing
Search for class

36. Protégé – Winter 2014
ClassesTab: Class Editor

37. Protégé – Winter 2014
ClassesTab: Class Editor
Class annotations (for class metadata)
Class name and documentation
Properties
“available”
to Class
Disjoints
widget
Conditions Widget
Class-specific tools (find usage etc)

38. Protégé – Winter 2014
Create Classes
1. Click the “Create Class” button
(this is above the class hierarchy)
A new class will be created as a subclass of owl:Thing
2. Type in a new name “DomainConcept” over the default
(return updates the hierarchy)
3. Req. for later labs: document your class using the rdfs:comment field
4. Create another class called “Pizza” using the same method
You will notice that Pizza has been created as a subclass of
DomainConcept as this was the class selected when the button was
pressed. You can also right-click any class and select “Create Class”
5. Create two more subclasses of DomainConcept called “PizzaTopping”
and “PizzaBase”.
Any mistakes, use the “Delete Class” button next to “Create Class”
Start with your empty ontology

39. Protégé – Winter 2014
Concept: Disjointness
All classes could potentially overlap
Pizza Ice Cream
= individual
This means an individual could be both a Pizza and a
IceCream at the same time.

40. Protégé – Winter 2014
Concept: Disjointness
If we state that classes are disjoint
This means an individual cannot be both a Pizza and a Ice
Cream at the same time
We must do this explicitly in the interface
Pizza Ice Cream
= individual

41. Protégé – Winter 2014
Concept: Disjointness
Add siblings as disjoint
Add new disjoint Remove disjoint siblings
List of disjoint classes

42. Protégé – Winter 2014
Classes
Create Pizza Ontology classes
 Pizza
 PizzaBase
 PizzaTopping

43. Protégé – Winter 2014
Classes
Pizza Toppings classes
Using the Tools->“Create class Hierarchy…”
 CheeseTopping
◦ MozzarellaTopping
◦ ParmezanTopping
 MeatTopping
◦ HamTopping
◦ PepperoniTopping
◦ SalamiTopping
◦ SpicyBeefTopping
 SeafoodTopping
◦ AnchovyTopping
◦ TunaTopping
 VegetableTopping
◦ OliveTopping
◦ OnionTopping
◦ PepperTopping
◦ TomatoTopping
Meaning: All individuals that are members of the class TomatoTopping
are members of the class VegetableTopping and members of the class
PizzaTopping…

44. Protégé – Winter 2014
Creating Properties
Properties can be in a hierarchy
Search for property
Create property

45. Protégé – Winter 2014
Creating Properties
There can be at most one individual (range)
that is related to the domain individual
via the property: “hasMother”
The inverse property is functional:
“hasMother → hasChild”
P(A,B) and P(B,C) -> P(A,C)
If a property P is transitive, and the property
relates individual A to individual B, and also
individual B to individual C: “hasAncestor”
P(A,B) -> P(B,A)
If a property P is symmetric, and the property
relates individual A to individual B : “hasSibling”

46. Protégé – Winter 2014
Properties in Protégé

47. Protégé – Winter 2014
Restrictions
Definition
 A restriction describes an anonymous
(unnamed) class
◦ The set of subclasses and individuals that satisfy
the given restriction
 Applies to
◦ Classes
◦ Object Properties
◦ Datatype Properties

48. Protégé – Winter 2014
Restrictions
Add restrictions
 Specify that Pizza only has one PizzaBase
◦ Add a cardinality restriction to Pizza on hasBase to exactly one
PizzaBase
◦ Add to the superclasses section “hasBase exactly 1 PizzaBase”
 Specify that Pizza has at least one PizzaTopping
◦ Add an existencial restriction to Pizza on hasTopping to
PizzaToppings
◦ Add to the superclasses section “hasTopping some PizzaTopping”
 Create a value restriction to specify that MozzarellaTopping
has Italy as its country of origin
◦ Create hasCountryOfOrigin object property with Country as range
◦ Add to the superclass section “hasCountryOfOrigin value Italy”

49. With Protégé 4.3
Using a Reasoner

50. Protégé – Winter 2014
Reasoner - Definition
 A piece of software able to infer logical
consequences from a set of asserted facts or
axioms

51. Protégé – Winter 2014
Reasoner -Used for
 Consistency checking
◦ Test whether a class could have instances
 Classification
◦ A classifier takes a class hierarchy and places a class in the class
hierarchy
◦ Task of turning implicit definitions already present in the
hierarchy as explicit

52. Protégé – Winter 2014
Reasoner- Protégé 4.3 support
 Built-in support
◦ Pellet (http://clarkparsia.com/pellet)
◦ FaCT ++(http://owl.man.ac.uk/factplusplus/)
 Connection to external reasoners through a DIG
interface

53. Protégé – Winter 2014
Reasoner - Pellet vs FaCT++
FaCT++Pellet
 Open-source Java OWL DL
reasoner
 Support expressivity of
SROIQ(D)
 Supports SWRL rules
 Available through AGPL
version 3 licence
 Latest version is 2.0.0 rc5
 Support in Protégé for 1.5
version
 New generation and C++
implementation of FaCT
 Support expressivity of
SROIQ(D)
 No support for Rules
 Available through GNU
public license
 Latest version is 1.2.3
 Support in Protégé for
1.2.3 version

54. Protégé – Winter 2014
 http://protegewiki.stanford.edu/index.php/Protege4UserDocs
 Pizzas and other ontologies are available from:
 www.co-ode.org/ontologies/
 W3C Documentation
 http://www.w3.org/2001/sw/BestPractices/
 Ian Horrocks, Sean Bechhofer:
 http://www.cs.man.ac.uk/~horrocks/Slides/Innsbruck-tutorial/
 Practical Guide To Building OWL Ontologies Using Prot´eg´e
4 and CO-ODE Tools, Mathew Horridge, Edition 1.2,
2009,university of manchester
Refrences

56. Protégé – Winter 2014
Pizzas…
Most often it is not the domain expert that formalises
their knowledge
Because of the complexity of the modelling task it is
normally a specialist “knowledge engineer”
 Hopefully, as tools get easier to use, this will change
Having access to experts is critical for most domains
Luckily, we are all experts in Pizzas, so we just need
some material to verify our knowledge…

57. Protégé – Winter 2014
Ontology in Linguistics
Concept
Form
/ Symbol
Referent
“semiotic
triangle”
activates Relates to
“Tank“
[Ogden, Richards, 1923]
Stands for
“a concept, is the
mediator that relates
the symbol to its
object” John F Sowa

58. Protégé – Winter 2014
Datatype Properties
Add datatype properties
 Create a hasCalorificContentValue datatype property with
integer as range
 Create a hasBaseProperty for PizzaBase (domain)
◦ Create a hasWidth datatype sub-property of hasBaseProperty with
integer as range
◦ Create a hasThickness datatype sub-property of hasBaseProperty
with float as range
 Change characteristics to Functional of the datatype
properties
◦ hasCalorificContentValue
◦ hasWidth
◦ hasThickness

59. Protégé – Winter 2014
Datatype Properties

60. Protégé – Winter 2014
Object Properties

61. Protégé – Winter 2014
Object Properties
Create some Object Properties
 isIngredientOf
◦ isToppingOf
◦ isBaseOf
 hasIngredient
◦ hasTopping
◦ hasBase
Pizza Ingredients
Ingredients that make up
the Pizza

62. Protégé – Winter 2014
Structure of an Ontology
Ontologies typically have two distinct components: Terminology
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 > 20 years
Background knowledge/constraints on the domain: Axioms
 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