INF3580 - V?r 2010

EXERCISES INF3580 SPRING 2010 WEEK 6

A bird.

A bird.

This document contains exercises made for INF3580. Please send any comments, errors, bug or improvement reports to this exercise set to martige@ifi.uio.no. Feedback is most welcome! Alphabetically thanks to Audun Stolpe, Espen H. Lian, Martin Giese and Rune Dahl for feedback.

The main curriculum for INF3580 spring 2010 is Semantic Web Programming by John Hebeler et al., Wiley Publishing, 2009. They have a website with additional articles and all source code used in the book at http://semwebprogramming.org/. Auxiliary curriculum is the book Foundations of Semantic Web Technologies by Hitzler, Krützsch, Rudolph, CRC Press 2009.

Keep all the work you do for these exercises in a safe place. Setting up a version control system like cvs, svn or git for the work you do is smart. You can create a svn repository on IfI's svn server, see their help section for more information. There is also a walk-through from old INF3120 on how to set up a svn repository and connect it to Eclipse, but news is that you'll need the plug-in subclipse to make it work. Please contact me if you have any smart tips to share.

1 RDFS

Read

  • Semantic Web Programming: chapter 4, 5.
  • Foundations of Semantic Web Technologies: chapter 2, 3.

1.1 RDFS vocabulary

We have already made an RDF representation of some of the members of the Simpson family. In this exercise we will use RDFS to define the classes and properties that we used to describe the Simpson family and also the classes and relationships that were missing from the vocabulary we were allowed to use in the RDF exercise.

We will also see how the added semantics for RDFS and the power of reasoners help us to discover implicit information.

Use RDFS to implement the following exercises. The syntax definition of RDFS is found at W3C's RDF Vocabulary Description Language 1.0: RDF Schema.

1.1.1 Exercise

Create a new RDF file. Add all work in this week's exercises to this file. Let fam be the prefix for the namespace 

http://www.ifi.uio.no/INF3580/v10/ex/family.n3#

1.1.2 Exercise

Create the classes and properties that we used in the RDF exercise, i.e., create the rdfs:Class-es

  • fam:Family
  • fam:Person

and the rdf:Property-es

  • fam:hasFamilyMember
  • fam:hasBrother
  • fam:hasSister
  • fam:hasParent
  • fam:hasMother
  • fam:hasFather
  • fam:hasSpouse
  • fam:hasName
1.1.2.1 Tip

This simply means that you shall state that fam:Family is an rdfs:Class and that fam:hasBrother is an rdf:Property: 

fam:Family rdf:type rdfs:Class .
fam:hasBrother rdf:type rdf:Property .

1.1.3 Exercise

Add more classes and properties.

Classes:

  • fam:Man
  • fam:Gender
  • fam:Woman

Properties:

  • fam:hasAunt
  • fam:hasChild
  • fam:hasDaugther
  • fam:hasGender
  • fam:hasGrandParent
  • fam:hasHusband
  • fam:hasSibling
  • fam:hasSon
  • fam:hasUncle
  • fam:hasWife
  • fam:hasAge

1.1.4 Exercise

State that fam:Female and fam:Male are of type fam:Gender.

1.1.5 Exercise

Add an rdf:Property fam:hasRelationshipTo which has fam:Person as both domain and range. This is intended as a general property which holds between two individuals related by any family relationship.

1.1.6 Exercise

Set the domain and range of fam:hasName to respectively fam:Person and xsd:string, and the domain and range of fam:hasAge to fam:Person and xsd:int respectively.

1.1.7 Exercise

For each class you now have created add the correct rdfs:subClassOf property assertions. For each property add the correct rdfs:subPropertyOf property assertions and the correct rdfs:domain and rdfs:range assertions.

For the model we have created we assume the "obvious" semantics, e.g., that the class Father is the class of all fathers and hasFather is the relationship from a person to its father. If you are in doubt, choose the interpretation you think is correct.

1.1.7.1 Tip

If we interpret the hasSibling property as a relationship between siblings and the hasSister property as a relationship from a sibling to a sister of the sibling, and interpret the classes Person and Woman in a way that fits this, then these relationships should look like the following in RDFS—this time in RDF/XML: 

<rdf:Property rdf:about="#hasSibling">
  <rdfs:domain rdf:resource="#Person"/>
  <rdfs:range rdf:resource="#Person"/>
</rdf:Property>
 
<rdf:Property rdf:about="#hasSister">
  <rdfs:subPropertyOf rdf:resource="#hasSibling"/>
  <rdfs:range rdf:resource="#Woman"/>
</rdf:Property>

1.2 The Simpson family

Now we will use the family vocabulary as a schema for the Simpsons data we have produced earlier, by opening both files in Protégé. Even though Protégé is an OWL editor it is quite safe to also load RDFS models as Protégé interprets them as OWL ontologies. This is not always the case, so it is best to use OWL if you do not need the meta-modelling capabilities of RDFS.

1.2.1 Exercise

Open Protégé and choose create a new OWL ontology. Give it the ontology URI 

http://www.ifi.uio.no/INF3580/v10/ex/simpsons.owl

Save it to a file of your choice, and choose the format you would link to use. What format you choose will not be directly visible in the user interface of Protégé, but is used when saving the ontology to file.

1.2.2 Exercise

Import the Simpsons RDF file you wrote in the RDF week exercises and the Family RDFS file you have written in this week's exercises. (This is done by clicking the plus sign in the "Ontology Imports" pane on the starting page of Protégé and importing "an ontology contained in a specific file" for each of the RDF files.) Note that Protégé seems to have a problem importing files which are not in RDF/XML format, while opening files in different formats works better. If you have written your files in a format different from RDF/XML and you have problems with this exercise, try converting your files to RDF/XML with, e.g., RDF Validator and Converter.

Then both files are successfully imported you should see the hierarchy of classes, properties and individuals under the tabs Classes, Object Properties and Data Properties respectively and Individuals. See also if your domain and range assertions look correct.

1.2.3 Exercise

Find the class Person in the Classes pane and see that it has quite a few members, while the classes Man and Woman have no members.

Now apply reasoning by choosing a reasoner, e.g., Pellet, in the Reasoner menu, and click Classify… in the same menu. Record any error messages that appear, you should get none.

If reasoning was successful, and assuming you have modelled classes and properties the same way as I have, you should see that the classes Man and Woman now have members.

Questions:

  • What is the added statements about Bart after reasoning?
  • In the Individuals list there might appear new individuals labelled genid1, genid2,… Explain what they are.
  • Which individuals do not get any added information after reasoning?

1.3 RDFS metrics

1.3.1 Exercise

Make a program which loads an RDF(S) file and outputs

  • the number of named rdfs:class-es,
  • the number of named rdfs:property-es,
  • the number of rdfs:domain assertions,
  • the number of rdfs:range assertions,
  • the number of rdfs:subClassOf axioms,
  • the number of rdfs:subPropertyOf axioms,
  • optionally, the maximum depth of the subclass hierarchy and
  • optionally, the maximum depth of the subproperty hierarchy.

The number of classes and properties should also include classes which are not explicitly declared as an rdfs:Class or rdf:Property (see Tip below), and not include classes or properties which are part of the RDF or RDFS vocabulary, e.g., rdf:type and rdfs:subClassOf.

The number of subclass and subproperty axioms should only include the axioms explicitly declared in the input file.

The maximum depth of the subclass hierarchy should count the maximum number of consecutive rdfs:subClassOf steps it is possible to make in the model, without stepping to an equivalent class. (If A is a subclass of B and B is subclass of A, then they are equivalent.) This means that you have to watch out for loops in the graph.

1.3.1.1 Tip

You should be able to make use of the RDF metrics program created in an earlier exercise. You will need to use an RDFS reasoner for some of the problems.

Running your metrics program on the following graph 

1:  @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
2:  @prefix : <http://example.org/> .
3:    :a a rdfs:Class ;
4:       rdfs:subClassOf :b .
5:    :b a rdfs:Class ;
6:       rdfs:subClassOf :c .

should give you an output similar to this: 

Named classes: 3
Named properties: 0
Domain axioms: 0
Range axioms: 0
Subclass axioms: 2
Subproperty axioms: 0
Max. depth of class tree: 2
Max. depth of property tree: 0

Note that even though :c is not explicitly typed as rdfs:Class, the class count returns 3. The depth of the subclass hierarchy is 2, since :a is a subclass of :b, which is a subclass of :c.

Running the following graph through your metrics program 

 1:  @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
 2:  @prefix : <http://example.org/> .
 3:    :a rdfs:subClassOf :b .
 4:    :b rdfs:subClassOf :c .
 5:    :c rdfs:subClassOf :d .
 6:    :d rdfs:subClassOf :b .
 7:    :c rdfs:subClassOf :e .
 8:    :e rdfs:subClassOf :f .
 9:    :f rdfs:subClassOf :g .
10:    :g rdfs:subClassOf :e .
11:  
12:    :a :relA :b .
13:    :a :relB :c .
14:  
15:    :relA rdfs:subPropertyOf :relB .
16:  
17:    :relA rdfs:range :b .
18:    :relB rdfs:domain :a .

should give you results similar to this: 

Named classes: 7
Named properties: 2
Domain axioms: 1
Range axioms: 1
Subclass axioms: 8
Subproperty axioms: 1
Max. depth of class tree: 2
Max. depth of property tree: 1

Note that there are two loops in the subclass hierarchy of this graph, :b – :c – :d – :b and :e – :f – :g – :e, which can "interfere" with the maximum depth calculation of the subclass hierarchy, if you are not careful.

/studier/emner/matnat/ifi/INF3580/v10/gruppeoppgaver/rdfsmetrics.png

Two loops.

1.3.2 Exercise

Find the metrics of your family RDFS file. Are the results as expected? Why / why not?

Author: Martin G. Skj?veland <martige@ifi.uio.no>

Date: 2010-03-12 15:42:16 CET

HTML generated by org-mode 6.34trans in emacs 23