Back to DBIS RDF Homepage
by Karsten Tolle (2000)
The Web
provides a simple and universal infrastructure to exchange various kinds of
information. In order to share, interpret, and manipulate information
worldwide, the role of metadata is widely recognized. Indeed, metadata allow us
to easily locate information available in the Web, by providing descriptions
about the structure and the content of the various Web resources (e.g. data,
documents, images, etc.) and for different purposes. The emergence of the
Resource Description Framework (RDF) is expected to enable metadata
interoperability across different communities or applications by supporting
common conventions about metadata syntax, structure, and semantics.
More
precisely, it provides a) a Standard Representation Language for Web metadata;
and b) a Schema Definition Language (RDFS) to interpret (meta)data using
specific class and property hierarchies (i.e. vocabularies). Moreover, RDF/RDFS
offer a syntax for representing metadata and schemas in XML, enabling the
creation and exchange of RDF descriptions in a both human readable and machine
understandable form. Many information providers like ABC News, CNN and Time
Inc., Web portals like Open Directory as well as Web browsers like Netscape,
and search engines like Altavista, Yahoo and Webcrawler already support the RDF
proposal.
RDF is
based on a directed graph model that alludes to the semantics of resource
description. The basic idea is that a Resource (identified by a URI) can be
described through a collection of Statements forming a so-called RDF
Description. A specific resource together with a named property and its value
is an RDF statement. RDFS schemas are then used to declare vocabularies, i.e.
collections of classes and properties, that can be used in resource
descriptions for a specific purpose or domain.
RDF
describes the interrelationships among resources in terms of named properties and values. These named properties may be thought of as attributes of
resources and in this sense correspond to traditional attribute-value pairs.
One main difference is that properties defined in RDF only identified by their
name (URI) and not like in other object models, where attributes are identified
by their name plus the domain class they can describe. So, properties have a
URI and therefore are also resources. The value of a property can be another resource or a literal. A literal is simple string or other primitive
data type as defined by [XML].
RDF
provides three representations of the RDF data model namely:
·
RDF
Graph – a syntax-neutral graphical description of the data
·
RDF
3-tuples – the set of statements described in triples
·
RDF
Syntax – provides some standard ways for describing data using XML
The RDF constructs and further nomenclature will be explained within the
section 0. For the other two models I just
explain the way of thinking and their underlying notion.
The RDF
Graph is a
syntax-neutral way of representing RDF expressions using directed labeled
graphs. These graphs also called nodes
and arcs diagrams.
The arcs represent the named properties. Each property connects two
nodes, coming from a node representing a resource (drawn as oval) and pointing
to another resource or a literal (drawn as rectangle). So, we have the
following two options, as shown in Figure 0.1:
Figure 0.1 Statements in the RDF Graph representation.
a)
The
named property p connects the two resources A and B.
b)
The
named property p connects the resource A with the literal L.
Such a
combination of a resource, a named property and its value is called a statement and define the resource as the subject, the property as the predicate and the value as the object of the
statement.
Figure 0.2 The elements of a statement: Subject, Predicate and Object.
This
statement would represent the following sentence:
Karsten
Tolle is the creator of the resource http://www.ics.forth.gr/proj/isst/RDF.
In some
cases we generate new resources to express a sentence. These new resources do
not have a predefined URI and we call them anonymous. Anonymous resources are represented in the RDF Graph as an empty oval.
Intern a URI will be created for this resource, otherwise it would not be one.
The internal URI is just not shown in this model and the choice how to generate
it has to be made by the application database designer. So, we can distinguish
between different anonymous resources. Consider as an example the following
sentence, Figure 0.3 represents this in graph form:
http://www.ics.forth.gr/proj/isst/RDF was created by an individual whose name is Karsten Tolle, email tolle@dbis.informatik.uni-frankfurt.de
and the individual whose name is Mister X published it.
Figure 0.3 Use of anonymous resources in RDF Graph.
I use the
‘dc’ prefix for the Dublin Core namespace as shown in Appendix B of [RDF
Schema] and the ‘s’ prefix for an imaginary namespace.
There are
many cases where it is necessary to make statements about a collection of
things, e.g., the members of a group or resources at a Web site. RDF provides
three different containers to make references to a collection of resources or
literal. A Bag representing a multi set of values, a Sequence representing an ordered list of values and an Alternative representing alternatives for a single value of a property. A container itself is a resource and must be instance of one of
the predefined classes (see RDF Schema for classes) rdf:Bag, rdf:Seq or rdf:Alt
by using the rdf:type property. The
properties to instanciate values to the container are named simply „rdf:_1“,
„rdf:_2“, „rdf:_3“, etc. For example, to represent the sentence
The
students of the course s:Digital Design are „Peter“, „Paul“ and „Mary“.
The
corresponding RDF Graph would be:
Figure 0.4 A bag as an example of how to use containers in RDF.
In order to enable RDF to make statements about other RDF statements it provides the ability to create a model of a statement. This process to make statements about statements is formally called reification in the Knowledge Representation community. So, a model of a statement is called a reified statement. A reified statement is a resource having the rdf:type = rdf:Statement and must define the following three properties to guarantee a unique reference:
- rdf:predicate The link to the predicate of the statement we refer to.
- rdf:subject The link to the subject of the statement we refer to.
- rdf:object The link to the object of the statement we refer to.
In Figure 0.5 you can see the abstract way of
modeling a statement and in Figure 0.6 we have a more precise example. In
both cases the resource for the reified statement is an anonymous resource. RDF
provides a shorthand to create statements about statements by using the
attribute ‘rdf:bagID’ (refer to [RDF M&S Chapter 4] for more examples).
Figure 0.5 The formal remodel of a statement to a reified statement.
Using a
reified statement we can model the sentence:
Vassilis says: ”Karsten is the creator of the resource http://www.ics.forth.gr/proj/isst/RDF.”
Figure 0.6 An example of a reified statement in the RDF Graph representation.
Another way
to represent a set of RDF statements is to use 3-tuples,
also called triples. Each triple {p, s,
o} correspond to an arc from the subject s to the object o, labeled by the
predicate p. The example shown in Figure 0.2 would be written as:
{[http://purl.org/dc/...#Creator], [http://www.ics.forth.gr/.../RDF], „Karsten Tolle“}
The
notation [I] denotes the resource
identified by the URI I and quotation
marks denote a literal. Actual this notation is only necessary for the object o
of the statement, because the predicate p and the subject s must be resources.
This
representation is an easy way to represent a very huge set of statements and
process them and therefore provides the scalability that had been forced as one
goal for RDF.
[RDF M&S] Ora Lassila and Ralph Swick: „Respirce Description
Framework (RDF) Model and Syntax Specification”
Online: http://www.w3.org/TR/REC-rdf-syntax
[RDF Schema] Brickley, D. ; Guha, R. V. : „Resource Description
Framework (RDF) Schema Specification“.
Online: http://www.w3.org/TR/rdf-schema
[XML] T. Bray, J.Paoli and C.M. Sperberg-McQeen: „Extensible Markup
Language (XML) 1.0“.
Online: http://www.w3.org/TR/REC-xml