usage of linked data - open university

Usage of Linked Data Introduction and Application Scenarios

Presented By:

Barry Norton

Agenda

1. Motivation Scenario

2. Linked Data Foundations

3. Introduction to Linked Data

4. Linked Data use case scenarios

2

MOTIVATION SCENARIO

3

Music!

4

• Provision of a music-based portal.

• Bring together a number of disparate components of data-oriented content:

1. Musical content (streaming data & downloads)

2. Music and artists metadata

3. Review content

4. Visual content (pictures of artists & albums)

Music!

5

Visualization Module

Metadata

Streaming providers

Physical Wrapper

Downloads

Dat

a ac

qu

isit

ion

D2R Transf. LD Wrapper

Musical Content

Ap

plic

atio

n

Analysis & Mining Module

LD D

atas

et

Acc

ess

LD Wrapper

RDF/ XML

Integrated Dataset

Interlinking Cleansing Vocabulary

Mapping

SPARQL Endpoint

Publishing

RDFa

Other content

Music!

6

Expected Results

• The developer will contribute back the aggregated and interlinked content to the Linked Open Data Cloud.

• Linking of artists will be improved.

• Metadata, visual content and reviews will be improved.

• Links to emerging Web technologies that inherit from semantics: Google RichSnippets, Facebook OpenGraph and schema.org annotation.

LINKED DATA FOUNDATIONS

7

• Extension of the technology of computer networks.

• The technology supporting the Internet includes the Internet Protocol (IP) .

• Each computer on the Internet is assigned an IP number.

• Messages can be routed from one computer to another.

Internet

8

Internet

9

Source: http://www.evolutionoftheweb.com

The growth of the Internet

• There is a wealth of information on the Web.

• It is aimed mostly towards consumption by humans as end-users: • Recognize the meaning behind content and draw conclusions,

• Infer new knowledge using context and

• Understand background information.

The Web

10

The Web

11

• Billions of diverse documents online, but it is not easily possible to automatically: • Retrieve relevant documents.

• Extract information.

• Combine information in a meaningful way.

• Idea: • Also publish machine processable data on the web.

• Formulate questions in terms understandable for a machine.

• Do this in a standardized way so machines can interoperate.

• The Web becomes a “Web of Data” • This provides a common framework to share knowledge on the Web

across application boundaries.

The Web: Evolution

12

Web of Documents Web of Data

“Documents”

Hyperlinks Typed Links

“Things”

The Web: Evolution

13

Source: http://www.radarnetworks.com

Web Technology Basics

14

HTTP

• Client server protocol.

• Every interaction based on:

Request Method

GET | PUSH | POST | PATCH | DELETE

(+ OPTIONS, HEADER, TRACE, CONNECT)

URL

Header

[Optional] Body (with POST, PUT, PATCH)

Response Response code (integer)

Header

[Optional] Body


15

HTTP Content Negotiation Example

GET /id/Barry_Norton HTTP/1.1 Host: www.aifb.kit.edu Accept:text/html

HTTP/1.0 302 Moved Temporarily Date: Sun, 07 Nov 2010 00:45:00 GMT Location: http://www.aifb.kit.edu/portal/index.php? title=Spezial:Exportiere_RDF/Barry_Norton

GET /id/Barry_Norton HTTP/1.1 Host: www.aifb.kit.edu Accept:application/rdf+xml

HTTP/1.0 302 Moved Temporarily Date: Sun, 07 Nov 2010 00:30:00 GMT Location: http://www.aifb.kit.edu/web/Barry_Norton

http://www.aifb.kit.edu/

http://www.aifb.kit.edu/


16

Uniform Resource Identifier (URI)

• Compact sequence of characters that identifies an abstract or physical resource.

• Examples:

ldap://[2001:db8::7]/c=GB?objectClass?one

mailto:[email protected]

news:comp.infosystems.www.servers.unix

tel:+1-816-555-1212

telnet://192.0.2.16:80/

urn:oasis:names:specification:docbook:dtd:xml:4.1.2

http://dbpedia.org/resource/Karlsruhe

Describing and Exchanging Data

17

Vocabularies

• Collections of URIs that can be used to represent information about a certain domain.

• Terms from well-known vocabularies should be reused wherever possible.

• New terms should be define only if you can not find required terms in existing vocabularies.


18

Vocabularies

A set of well-known vocabularies has evolved in the Semantic Web community. Some of them are:

Vocabulary Description

Friend-of-a-Friend (FOAF) Vocabulary for describing people.

Dublin Core (DC) Defines general metadata attributes.

Semantically-Interlinked Online Communities (SIOC)

Vocabulary for representing online communities.

Music Ontology (MO) Provides terms for describing artists, albums and tracks.

Simple Knowledge Organization System (SKOS)

Vocabulary for representing taxonomies and loosely structured knowledge.


19

Vocabularies

More extensive lists of well-known vocabularies are maintained by:

• W3C SWEO Linking Open Data community project http://www.w3.org/wiki/TaskForces/CommunityProjects/LinkingOpenData/CommonVocabularies

• Mondeca: Linked Open Vocabularies http://labs.mondeca.com/dataset/lov

• Library Linked Data Incubator Group: Vocabularies in the library domain

http://www.w3.org/2005/Incubator/lld/XGR-lld-vocabdataset-20111025


20

Extensible Markup Language (XML)

• Markup language that defines a set of rules for encoding documents.

• Main structure: tag.

• Human-readable and machine-readable.

– Every piece of information is self-described.

– Relations are also defined through the nesting structure.

• Allows the definition of constraints on values.


21


XML Element:

• Represents “things”.

• Consists of: opening tag, content, closing tag. <band>The Beatles</band>

• Content may be text, other elements or nothing.

• If there is no content, the element is empty and it is abbreviated as follows:

<band/> for <band></band>


22

Extensible Markup Language (XML) (Example)

<?xml version="1.0" encoding="UTF-8"?>

<musicArtists>

<band>

<name>The Beatles</name>

<genre>Rock</genre>

<member>John Lennon</member>

<member>Paul McCartney</member>

<member>George Harrison</member>

<member>Ringo Starr</member>

</band>

</musicArtists>

XML declaration

Opening tag

Closing tag

Content: - Other elements - Text - Nothing


23


XML Attribute:

• Represents a property of the “thing”.

• Name-value pair inside the opening tag of an element. <band name=“The Beatles”>...</band>

• Attributes can be replaced by elements.

• Attributes cannot be nested.

Semantics on the Web

24

Semantic Web Stack Berners-Lee (2006)


25


Syntatic basis

Basic data model

Simple ontology language

Expressive language

Query language

Application specific declarative-knowledge

Digital signatures, recommendations

Proof generation, exchange, validation


26


RDF


27

RDF

• RDF is the basis layer of the Semantic Web stack ‘layer cake’.

• Basic building block: RDF triple.

• Subject – a resource, which may be identified with a URI.

• Object – a resource or literal to which the subject is related.

• Predicate – a URI-identified reused specification of the relationship.


28

RDF (Example)

<http://musicbrainz.org/artist/b10bbbfc-

cf9e-42e0-be17-e2c3e1d2600d>

<http://www.w3.org/2002/07/owl#sameAs>

<http://dbpedia.org/resource/The_Beatles>.

<http://musicbrainz.org/artist/b10bbbfc-

cf9e-42e0-be17-e2c3e1d2600d>

<http://xmlns.com/foaf/0.1/name>

“The Beatles”.

URIs are given in angle brackets.

Literals are given in quotes.


29

RDF Graphs

• Every set of RDF assertions can then be drawn and manipulated as a (labelled directed) graph: • Resources – the subjects and objects are nodes of the graph.

• Predicates – each predicate use becomes a label for an arc, connecting the subject to the object.

Subject Object Predicate

30

<http://musicbrainz.org/artist/b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d>

<http://www.w3.org/2002/07/owl#sameAs>


<http://dbpedia.org/resource/The_Beatles> “The Beatles”


RDF Graphs (Example)


31

RDF Blank Nodes • RDF graphs can also contain unidentified resources, called

blank nodes:

• Blank nodes can group related information, but their use in Linked Data is discouraged.

<http://musicbrainz.org/artist/b10bbbfc- cf9e-42e0-be17-e2c3e1d2600d>

<http://music.owl/founded>

[]

“Apple Records”



32

RDF Turtle

• Turtle is a syntax for RDF more readable.

• Since many URIs share same basis we use prefixes: @prefix rdfs:<http://www.w3.org/2000/01/rdf-schema#>.

@prefix owl:<http://www.w3.org/2002/07/owl#>.

@prefix mbz:<http://musicbrainz.org/artist/>.

@prefix mo:<http://purl.org/ontology/mo/>.

@prefix dbpedia:<http://dbpedia.org/resouce/>.

@prefix bbc:<http://www.bbc.co.uk/music/artists/>.

• A simple shorthand is for class membership: mbz:b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d a mo:MusicGroup .

<http://www.w3.org/1999/02/22-rdf-syntax-ns#type>

Same subject & predicate


33

RDF Turtle

• When multiple statements apply to same subject they can be abbreviated as follows:

mbz:b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d rdfs:label “The Beatles”;

owl:sameAs dbpedia:The_Beatles ,

bbc:b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d .

• Turtle also provides a simple syntax for datatypes and language tags for literals, respectively:

dbpedia:The_Beatles foaf:name “The Beatles”^^xsd:string; dbpedia-owl:abstract “The Beatles were an English rock band formed (…) ”@en,

“The Beatles waren eine britische Rockband in den (…) ”@de.

Same subject


34

RDF/XML

• This is most useful for inter-machine communication (due particularly to programming support for XML).

• As XML, an RDF/XML document should start with an XML version and encoding:

<?xml version="1.0" encoding="utf-8"?>

• The root element is then as follows:

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"> ... </rdf:RDF>


35

RDF/XML

• The primary (recurring) element in encoding assertions (thereby triples) is rdf:Description, e.g.:

<rdf:Description

rdf:about="http://musicbrainz.org/artist/b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d">

<foaf:name>The Beatles</foaf:name>

<owl:sameAs rdf:resource="http://dbpedia.org/resource/The_Beatles">

</rdf:Description>

<rdf:Description

rdf:about="http://musicbrainz.org/artist/4d5447d7-c61c-4120-ba1b-d7f471d385b9">

<foaf:name>John Lennon</foaf:name>

</rdf:Description>


36


RDF-S


37

RDF Schema (RDFS)

Language for two tasks w.r.t. the RDF data model:

• Expectation – nominate:

• the ‘types’, i.e., classes, of things we might make assertions about, and

• the properties we might apply, as predicates in these assertions, to capture their relationships.

• Inference – given a set of assertions, using these classes and properties, specify what should be inferred about assertions that are implicitly made.


38

RDF Schema (RDFS)

Resources and predicates with (limited) inferences: rdfs:Resource

rdfs:Literal, rdfs:DataType

rdfs:Class, rdfs:subClassOf

rdfs:subPropertyOf

rdfs:range, rdfs:domain

Some predicates with NO inferences: rdfs:comment

rdfs:label

rdfs:seeAlso

rdfs:isDefinedBy


39

RDF Schema (RDFS)

dbpedia:Britpop rdfs:subClassOf dbpedia:Rock . dbpedia:The_Beatles rdf:type dbpedia:Britpop . dbpedia:The_Beatles rdf:type dbpedia:Rock .

Schema

Existing fact

Inferred fact

We expect to use this vocabulary to make assertions about music genres.

Having made such an assertion...

Inferences can be drawn that we did not explicitly make


40

RDF Schema (RDFS)

From schema. From properties. From sub-properties.

Inferences


41


Ontology: OWL


42

Ontology: OWL

• RDFS provides a simplified ontological language for defining vocabularies about specific domains.

• Sometimes it is necessary to have access to a wider range of ontological constructs.

• Web Ontology Language (OWL) provides more ontological constructs and avoids some of the potential confusion in RDFS.


43

Ontology: OWL

• OWL extends existing Web standards (XML, RDF model).

• Own namespace and own terms in the namespace.

• OWL has roots in Description Logics: reuse of reasoners.

• OWL formal semantics:

• Based on a direct model theory for OWL.

• Compatible with existing Description Logics and can be mapped to them.

• User-friendly and high-level syntax which is then mapped to RDF.


44

Ontology: OWL In order to capture different requirements towards scalability and expresivity, OWL consists of different language variants.

Full

DL

Lite

OWL Full • Union of OWL syntax and RDF/S. • Includes the full expressivity of RDF(S) and

is therefore computationally intractable.

OWL DL • Restricted to a subset conforming to

Description Logics and is therefore computationally tractable.

OWL Lite • Simplified subset of OWL DL.


45

Ontology: OWL DL

• We will focus on the OWL DL language: • OWL Lite is simply a subset of this.

• OWL Full removes the remaining restrictions, but typically results in computationally intractable ontologies.

• OWL is made up of terms which provide for: • Class construction.

• Property construction.

• Class axioms.

• Property axioms.

• Individual axioms.


46

Ontology: OWL DL – Class Construction These class constructs are available in OWL, not in RDFS .

• The class of pop-rock music genre. intersectionOf (Pop, Rock)

• The class of music artists. unionOf (SingleArtist, MusicBand)

• Everything that’s not instrumental music (Not the same as the class vocal music).

complementOf (InstrumentalMusic)


47

Ontology: OWL DL – Property Construction

OWL distinguishes between OWL ObjectProperties (resources as values) and OWL DatatypeProperties (literals as values).

:plays rdf:type owl:ObjectProperty; rdfs:domain :Musician; rdfs:range :Instrument . :hasMembers rdf:type owl:DatatypeProperty; rdfs:range xsd:int .


48

Ontology: OWL DL – Class Axioms Axioms permit us to make general statements about concepts which are used in logical inference (reasoning). Class axioms:

• Sub-class relationship (from RDF Schema). subClassOf (Man, Human)

• Equivalent relationship (classes have the same individuals).

equivalentClass (Man, intersectionOf(Male, Human))

• Disjointness (classes have no shared individuals).

disjointClasses (Man, Woman)


49

Ontology: OWL DL – Property Axioms OWL also allows for property equivalence and disjointness:

• If somebody is an university employee, then they work for the university (or vice versa).

equivalentProperty(isUniversityEmployee, worksForUniversity)

• A music band is not a single artist. propertyDisjointWith(isBand, isSingleArtist)


50

Ontology: OWL DL – Property Axioms OWL introduces property characteristics for more expressivity in inferring characteristics relating to instances and their properties.

• Transitivity (if X p Y and Y p Z then X p Z)

• Symmetry (if X p Y then Y p X)

• Functional (if X p Y and X p Z then Y = Z)

• Inverse (if X p1 Y then Y p2 X)


51

Ontology: OWL DL – Individual Axioms OWL Individuals represent instances of classes. They are related to their class by the rdf:type property.

• It’s possible to explicitly state that two individuals are the same.

sameAs(Sir James Paul McCartney, Paul McCartney)

• It’s possible to explicitly state that two individuals are different.

differentFrom (TheBeatles_band, TheBeatles_TVseries)


52

Ontology: OWL DL – Cardinalities

A cardinality is a specification of how many different values can be given to a property on an individual of a particular class.

• The class of bands with 4 or more members. minCardinality(hasMember, 4)

• The class of bands with no more than 3 members. maxCardinality(hasMember, 3)

• The class of bands with exactly 2 members. maxCardinality(hasMember, 2)


53


Query: SPARQL


54

Query: SPARQL

• Query language designed to use a syntax similar to SQL for retrieving data from relational databases.

– SELECT nominates which components of the matches made against the data should be returned.

– FROM (optional) indicates the sources for the data against which to find matches.

– WHERE defines patterns to match against the data.

– ORDER BY defines a means to order the selected matches.


55

Query: SPARQL (Example)

PREFIX dc: <http://purl.org/dc/elements/1.1/> PREFIX foaf: <http://xmlns.com/foaf/0.1/> PREFIX dbpedia: <http://dbpedia.org/resource/> PREFIX music-ont: <http://purl.org/ontology/mo/>

SELECT ?album_name ?track_title WHERE { <http://musicbrainz.org/artist/b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d> foaf:made ?album . ?album dc:title ?album_name . ?album music-ont:track ?track . ?track dc:title ?track_title . }

Retrieve the names of the albums and tracks recorded by The Beatles.

http://purl.org/dc/elements/1.1/

http://xmlns.com/foaf/0.1/

http://dbpedia.org/resource/

http://purl.org/ontology/mo/

http://musicbrainz.org/artist/b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d











56

Query: SPARQL SQL SPARQL

Based on relations (tables). Based on labelled directed graphs.

The relations (tables) to be matched over should be indicated.

Assumes a default graph. (The FROM clausde indicates further named graphs).

(Retrieval) queries produce a relation from a relation.

SPARQL SELECT queries produce a relation from a graph. CONSTRUCT queries (considered later) produce a graph from a graph.


57

Query: SPARQL As well as triple patterns the following are graph patterns:

• OPTIONAL {graph pattern} Graph patterns that only bind variables if they are matched.

• {graph pattern 1} UNION {graph pattern 2} Alternative matches to bind variables (one or the other pattern must match).

• FILTER (condition) Allows boolean conditions (defined using a number of built-in functions – especially over datatyped literals) to preempt certain matches.


58

Query: SPARQL After the WHERE clause, SPARQL allows for some qualifiers which further process the results set before it is returned.

• ORDER BY expression Controls the order of the results according to the expression.

• LIMIT possitive-integer Limits the number of results returned to the integer given.

• OFFSET possitive-integer Returns only the results in the result list after the given integer, i.e. if the integer is 5, then only results from the 5th on are given.

INTRODUCTION TO: LINKED DATA

59

Linked Data

60

• Set of best practices for publishing data on the Web.

• Data from different knowledge domains, self-described, linked and accesible.

• Follows 5 simple principles…

Linked Data Principles

1. Use URIs as names for things.

2. Use HTTP URIs so that users can look up those names.

3. When someone looks up a URI, provide useful information, using the standards (RDF*, SPARQL).

4. Include links to other URIs, so that users can discover more things.

61


• A foundational issue in Linked Data was the distinction of URIs for real-world objects documents that might describe them.

62

1. Use URIs as names for things.


• HTTP allows a second way to distinguish real-world objects from documents, e.g. In DBPedia:

• Principles say HTTP 303 and Location header should be used.

63

2. Use HTTP URIs so user can look up those names.


64


• While RDF/XML should be the default for look-up.

• RDFa annotations in HTML are now also standard.

• A dump of the whole dataset, and a SPARQL endpoint for queries are also encouraged.


65


What to return for a URI?

• Immediate description: triples where the URI is the subject.

• Backlinks: triples where the URI is the object.

• Related descriptions: information of interest in typical usage scenarios.

• Metadata: information as author and licensing information.

• Syntax: RDF descriptions as RDF/XML and human-readable formats.

There are several ways to reuse URIs: • direct reuse of class/property

• (RDFS) sub-class/-property

• (OWL) equivalent class/property

• SKOS broad match

• direct reuse

• (RDFS) seeAlso

• (OWL) sameAs


66

4. Include links to other URIs, so that users can discover more things.

Schema Level

Instance Level

Linked Data 5 Star

67

Data is available on the Web.

Data is available as machine-readable structured data.

Non-propietary formats are used.

Individual data identified with open standards.

Data is linked to other data provider.

Linked Data 5 Star

68

Example:

„John Lennon“

„Paul McCartney“

„George Harrison“

„Ringo Starr“

My Data

Linked Data 5 Star

69

Data is available on the Web

„John Lennon“



„Ringo Starr“

My Data

It can be accesed, downloaded…

Linked Data 5 Star

70

Data is available as machine-readable structured data

„John Lennon“



„Ringo Starr“

My Data

„The Beatles“ http://upload.wikimedia.org/wikipedia/commons/thumb/d/df/The_Fabs.JPG/600px-The_Fabs.JPG

Please Please Me – 1963 A Hard Day‘s Night – 1964 Help! – 1965 Revolver – 1966 …

Images Scanned Information Plain text

Machine-readable data:

Linked Data 5 Star

71

Non-propietary formats are used

My Data

<band> <name>The Beatles</name> <member>John Lennon</member> <member>Paul McCartney</member> <member>George Harrison</member> <member>Ringo Starr</member> <picture>http://upload.wikimedia.org/wikipedia/commons/thumb/d/df/The_Fabs.JPG/600px-The_Fabs.JPG</picture> <album year=1963>Please Please Me</album> <album year=1964>A Hard Day‘s Night</album> <album year=1965>Help!</album> <album year=1966>Revolver</album> … </band>

Linked Data 5 Star

72

Individual data identified with open standards

<band> <name>http://musicbrainz.org/artist/b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d</name> <member>http://musicbrainz.org/artist/4d5447d7-c61c-4120-ba1b-d7f471d385b9</member> ... <album year=1963>http://musicbrainz.org/release/5f3ba07b-4a24-4cd5-b8ad-95ba0fcebec1</album> … </band>

My Data

In this context, „Revolver“ is an album! Not a gun.

URI: Uniform Resource Identifier • Data is univocacally identified

The Beatles John Lennon

Revolver • Dissambiguation


























http://musicbrainz.org/artist/4d5447d7-c61c-4120-ba1b-d7f471d385b9


























http://musicbrainz.org/release/5f3ba07b-4a24-4cd5-b8ad-95ba0fcebec1

























Linked Data 5 Star

73

Data is linked to other data provider

<band> <name>http://musicbrainz.org/artist/b10bbbfc-cf9e-42e0-be17-e2c3e1d2600d</name> <member>http://musicbrainz.org/artist/4d5447d7-c61c-4120-ba1b-d7f471d385b9</member> ... <album year=1963>http://musicbrainz.org/release/5f3ba07b-4a24-4cd5-b8ad-95ba0fcebec1</album> … </band>

http://dbpedia.org/resource/John_Lennon






























Linked Data Cloud

74

2007

Linked Data Cloud

75

2008

Linked Data Cloud

76

2009

Linked Data Cloud

77

2010

Linked Data Cloud

78

2011

Linked Data Cloud

79

2012

State of the LOD Cloud1

80

• Total Datasets:

295

• Total Triples:

31,634,213,770

• Total (Out-)Links:

503,998,829

1 Version 0.3, 09/19/2011 http://www4.wiwiss.fu-berlin.de/lodcloud/state

http://www4.wiwiss.fu-berlin.de/lodcloud/state





Exploring the Web of Data

81

• Linked Data browsers

• Linked Data mashups

• Search engines

Linked Data Browsers

82

http://marbles.sourceforge.net

Marbles

http://marbles.sourceforge.net/

Linked Data Mashup

83

http://revyu.com

Revyu.com

http://revyu.com/

Linked Data Mashup

84

DBPedia Mobile

Pictures from revyu.com

http://wiki.dbpedia.org/DBPediaMobile

http://wiki.dbpedia.org/DBPediaMobile

Linked Data Mashup

85

http://sig.ma

SIGMA

http://sig.ma/

Linked Data Search Engines

86

http://iws.seu.edu.cn/services/falcons

Falcons



Linked Data Search Engines

87

http://data.nytimes.com/schools/schools.html

NYTimes



Some Application Scenarios

88

BBC


89

LinkedGeoData.org


90

Linked Government Data: USA


91

Linked Government Data: UK

Summary

92

In this chapter we studied:

• The Web and its evolution.

• Web technology basics: HTTP, URI.

• Vocabularies and XML to describe and exchange data.

• The Semantic Web stack: RDF, RDFS, OWL, SPARQL.

• Linked Data concept and principles.

• Evolution of the LOD cloud.

• Browsers, mashups and search engines to explore the Web of Data.

• Some application scenarios.

93

@euclid_project