Databases Reference
In-Depth Information
Turtle.
Turtle (Terse RDF Triple Language) is a subset of, and compatible with, Nota-
tion 3 and a superset of the minimal N-Triples format (cf. Figure 4). The goal was to use
the essential parts of Notation 3 for the serialization of RDF models and omit everything
else. Turtle became part of the SPARQL query language for expressing graph patterns.
Compared to N-Triples, Turtle introduces a number of shortcuts, such as namespace
definitions (lines 1-5 in Figure 5), the semicolon as a separator between triples sharing
the same subject (which then does not have to be repeated in subsequent triples) and
the comma as a separator between triples sharing the same subject and predicate. Tur-
tle, just like Notation 3, is human-readable, and can handle the "%" character in URIs
(required for encoding special characters) as well as IRIs due to its UTF-8 encoding.
Notation 3.
N3 (Notation 3) was devised by Tim Berners-Lee and developed for the
purpose of serializing RDF. The main aim was to create a very human-readable serial-
ization. Hence, an RDF model serialized in N3 is much more compact than the same
model in RDF
XML but still allows a great deal of expressiveness even going beyond
the RDF data model in some aspects. Since, the encoding for N3 files is UTF-8 the use
of IRIs does not pose a problem.
/
RDF
XML syntax [98] is standardized by the W3C and is widely used
to publish Linked Data on the Web. However, the syntax is also viewed as di
/
XML.
The RDF
/
cult
for humans to read and write, and therefore consideration should be given to using
other serializations in data management and curation workflows that involve human
intervention, and to the provision of alternative serializations for consumers who may
wish to eyeball the data. The MIME type that should be used for RDF
/
XML within
HTTP content negotiation is application/rdf+xml.
RDFa.
RDF in Attributes (RDFa, [2]) was developed for embedding RDF into XHTML
pages. Since it is an extension to the XML based XHTML, UTF-8 and UTF-16 are
used for encoding. The "%" character for URIs in triples can be used because RDFa
tags are not used for a part of a RDF statement. Thus IRIs are usable, too. Because
RDFa is embedded in XHTML, the overhead is higher compared to other serialization
technologies and also reduces the readability. The basic idea of RDFa is enable an RDFa
processor to extract RDF statements from an RDFa enriched HTML document. This is
achieved by defining the scope of a certain resource description, for example, using the
'about' attribute (cf. line 10 in Figure 5). Within this scope, triples can now be extracted
from links having an additional 'rel' attribute (line 13) or other tags having a 'property
attribute' (lines 11 and 14).
JSON-LD.
JavaScript Object Notation (JSON) was developed for easy data interchange
between applications. JSON, although carrying JavaScript in its name and being a
subset of JavaScript, meanwhile became a language independent format which can be
used for exchanging all kinds of data structures and is widely supported in di
erent pro-
gramming languages. Compared to XML, JSON-LD requires less overhead with regard
to parsing and serializing. JSON-LD has been developed by the JSON for Linking Data
Community Group and been transferred to the RDF Working Group for review, im-
provement, and publication along the Recommendation track. JSON-LD's design goals
ff
