Information Technology Reference
In-Depth Information
formats abstract from concrete domains and represent different types of knowledge
in a unified graph aggregating heterogeneous data sources and knowledge types.
5.3.1 Semantic Data Formats
Semantic techniques for representing knowledge are based on ontologies and graphs.
Ontologies are designed to store various types of knowledge in a unified machine
readableway. This enablesmachines to understand themeaning of data and simplifies
the sharing and reuse of data in different scenarios [ 14 ].
The semantic knowledge representation is based on ontologies defining the rel-
evant aspects of the modeled domain. Ontologies define the concepts and the rele-
vant relationships between the concepts. In addition, ontologies may contain rules
enabling deriving implicit knowledge as well as checking the consistency of knowl-
edge. Ontologies describe the structure of the domain and define the basic concepts.
Knowledge about the entities and the relationship between entities is represented
based on graphs consisting of nodes and edges. The nodes describe concepts and
instances, the edges the relations between the nodes. In general, the edges are labeled
allowing a fine-grained description of the relations. In order to store graphs in a flat
way, graphs can be decomposed into a set of triplets consisting of subject, predicate,
and object.
The most popular data formats used for storing semantic knowledge are the Web
Ontology Language (OWL) [ 11 ] and the Resource Description Framework (RDF)
[ 3 , 4 ]. Both data formats are endorsed by the W3C and enable the efficient repre-
sentation of semantic knowledge. RDF focusses on the representation of knowledge
as triples, OWL supports additionally semantic relations (e.g., sameAs ) and log-
ical predicate logic (reasoning). Using RDF and OWL, comprehensive knowledge
resources can be built providing a valuable knowledge base for a wide variety of
scenarios.
5.3.2 Semantic Resources
Knowledge resources providing semantically represented knowledge exists for many
domains [ 6 ]. The most popular knowledge resources are visualized in the Linked-
Open-Data Cloud (see Fig. 5.1 ).
The central node of the linked open data cloud is the DBpedia [ 2 ]. DBpedia
contains data extracted from Wikipedia in a semantic data format (RDF). DBpedia
provides knowledge for a wide variety of domains and acts as the most important
hub for connecting the different sources in the Linked Open Data cloud.
Search WWH ::




Custom Search