Menu
Home
Log in / Register
 
Home arrow Computer Science arrow Linked Open Data
< Prev   CONTENTS   Next >

2.3 RDF Data Model

The RDF data model [7] represents information as sets of statements, which can be visualized as node-and-arc-labeled directed graphs. The data model is designed for the integrated representation of information that originates from multiple sources, is heterogeneously structured, and is represented using different schemata. RDF can be viewed as a lingua franca, capable of moderating between other data models that are used on the Web.

In RDF, information is represented in statements, called RDF triples. The three parts of each triple are called its subject, predicate, and object. A triple mimics the basic structure of a simple sentence, such as for example:

Burkhard Jung

is the mayor of

Leipzig

(subject)

(predicate)

(object)

The following is the formal definition of RDF triples as it can be found in the W3C RDF standard [7].

Definition 1 (RDF Triple). Assume there are pairwise disjoint infinite sets I, B, and L representing IRIs, blank nodes, and RDF literals, respectively. A triple (v1, v2, v3) (I ∪ B) × I × (I ∪ B ∪ L) is called an RDF triple. In this tuple, v1 is the subject, v2 the predicate and v3 the object. We call T = I ∪ B ∪ L the set of RDF terms.

The main idea is to use IRIs as identifiers for entities in the subject, predicate and object positions in a triple. Data values can be represented in the object position as literals. Furthermore, the RDF data model also allows in subject and object positions the use of identifiers for unnamed entities (called blank nodes), which are not globally unique and can thus only be referenced locally. However, the use of blank nodes is discouraged in the Linked Data context. Our example fact sentence about Leipzig's mayor would now look as follows:

<leipzig.de/id>

<example.org/p/hasMayor>

<Burkhard-Jung.de/id> .

(subject) (predicate) (object)

This example shows that IRIs used within a triple can originate from different namespaces thus effectively facilitating the mixing and mashing of different RDF vocabularies and entities from different Linked Data knowledge bases. A triple having identifiers from different knowledge bases at subject and object position can be also viewed as an typed link between the entities identified by subject and object. The predicate then identifies the type of link. If we combine different triples we obtain an RDF graph.

Definition 2 (RDF Graph). A finite set of RDF triples is called RDF graph. The RDF graph itself represents an resource, which is located at a certain location on the Web and thus has an associated IRI, the graph IRI.

Fig. 3. Example RDF graph describing the city of Leipzig and its mayor.

An example of an RDF graph is depicted in Fig. 3. Each unique subject or object contained in the graph is visualized as a node (i.e. oval for resources and rectangle for literals). Predicates are visualized as labeled arcs connecting the respective nodes. There are a number of synonyms being used for RDF graphs, all meaning essentially the same but stressing different aspects of an RDF graph, such as RDF document (file perspective), knowledge base (collection of facts), vocabulary (shared terminology), ontology (shared logical conceptualization).

2.4 RDF Serializations

The initial official W3C RDF standard [7] comprised a serialization of the RDF data model in XML called RDF/XML. Its rationale was to integrate RDF with the existing XML standard, so it could be used smoothly in conjunction with the existing XML technology landscape. However, RDF/XML turned out to be difficult to understand for the majority of potential users because it requires to be familiar with two data models (i.e., the tree-oriented XML data model as well as the statement oriented RDF datamodel) and interactions between them, since RDF statements are represented in XML. As a consequence, with NTriples, Turtle and N3 a family of alternative text-based RDF serializations was developed, whose members have the same origin, but balance differently between readability for humans and machines. Later in 2009, RDFa (RDF Annotations, [1]) was standardized by the W3C in order to simplify the integration of HTML and RDF and to allow the joint representation of structured and unstructured content within a single source HTML document. Another RDF serialization, which is particularly beneficial in the context of JavaScript web applications and mashups is the serialization of RDF in JSON. Figure 4 presents an example serialized in the most popular serializations.

 
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Philosophy
Political science
Psychology
Religion
Sociology
Travel