This paper outlines the Telefónica's position on the Future of RDF.
RDF is world-wide recognized as the standard representation format for linked data published on the Web. In fact the RDF's resource triple model has been proved to be flexible enough to interconnect heterogenous and distributed data represented using multiple vocabularies. Nonetheless, the current (2004) version of RDF [RDF-CONCEPTS] presents several caveats which, in our opinion, call for further design and standardization actions. Hence, this position paper describes what are, from our point of view, the most important aspects on which action needs to be taken, specially in the context of the Open Linked Data initiative.
The current technology for specifying RDF vocabularies is RDF Schema [RDF-SCHEMA], which uses RDF itself to describe vocabularies. Around RDF Schema we envisage the following enhancements:
In our opinion the absence of an easy-to-be-read syntax has been one of the major stoppers for the adoption of RDF as a mainstream technology. RDF/XML [RDF-SYNTAX], the only syntax Recommended by W3C, presents different issues. On one hand, although it is based on XML, RDF/XML is not intended to be human readable. Furthermore, sometimes when a mainstream developer looks at the RDF/XML syntax she feels overwhelmed and eventually gives up. On the other hand, the RDF/XML syntax is very verbose. This could be an issue for the computational infrastructure (for instance, triple-stores) devoted to support Open Linked Data.
N3, NTriples and Turtle [TURTLE] are intended to be an alternative to RDF/XML. Nevertheless, from our point of view, these notations try to solve the problems of human readability and compactness at the same time. In addition they are plain text formats that have to be processed using specific parsers. Thus, to tackle the RDF/XML problems described above we propose the following work:
A number of properties in an RDF Vocabulary can be used to represent physical magnitudes. The same magnitude, for instance "length", can be measured using different units ("meters", "inches", "feet", "milimeters", etc), yielding to different numeric values depending on the unit used. However, from a conceptual point of view, all of these values are (or should be) actually equivalent. RDF (nor OWL) supports to tag a literal value with its measurement unit. In other words, RDF literal nodes can represent discrete numeric values, such as "10" or "9.81", but a literal node does not capture the unit used to express the value.
A simple solution is to express values normatively in a single unit, for example the one mandated by the International System of Units (SI). However, it is also recognized that this approach might not be suitable in all scenarios, due to both practical (some scales are more conveniently captured using certain units) and cultural reasons. Thus, it is needed a more general solution to publish units-aware Open Linked Data on the Web. Thus, We propose to use the Measurement Units Ontology (MUO) [MUO] which it is a result coming from the Eureka-Celtic MyMobileWeb [MYMW] project. In fact this Ontology was developed as an extension to the W3C's Delivery Context Ontology [DCO].
In order to understand how the MUO Ontology works, we use the example depicted below, which it is an RDF representation of the area of Spain (504,782 Square Kilometres). It can be seen that the property
area (a physical magnitude) has been modelled as a subproperty of
muo:qualityValue. The range of such a Property is a resource of type
muo:QualityValue with has two properties: the numerical area value and the measurement unit on which such a value is expressed. As it can observed, the unit (
muoinst:Sq_Km) is expressed as an instance of the
Figure 1 .- MUO usage example (the area of Spain)
Hence, We propose to incorporate the MUO design patterns and properties as part of the set of standard RDF Properties and classes, thus enabling the creation of units-aware RDF Graphs. This advance will be of extremely importance to the Open Linked Data initiative as a big amount of data (specially government data) will need to published along with its measurement units.
XML is being pervasively used by businesses and developers as a data representation format. Web believe that standard interoperability mechanisms to systematically convert XML documents into RDF Graphs are needed for the following reasons: (a) to estimulate the publication of data on the Web and (b) to foster the adoption of RDF as a data representation format.
One of the reasons of the success of XML has been the availability of a language and technology independent API, the DOM. The DOM is intended to create, parse or manipulate data represented in XML. Unfortunately there is no a DOM for RDF graphs. There are different APIs in different languages, thus developers need to constantly learn new data structures, functions or patterns. Hence, we believe it is urgent to standardize a language-independent and generic API for the manipulation (including querying) of RDF graphs ("the RDFOM"). The RDFOM should be based on the experience and knowledge gained from the design of existing APIs for RDF (Jena [JENA], Redland [REDLAND], ...). At this respect, and taking into account the DOM experience, we believe that simplicity should be an important design principle for the RDFOM, thus convenience functions should be provided when possible.
Telefónica believes that RDF should be enhanced in order to face with the new challenges posed by the Open Linked Data movement. To this aim we have identified different work items to be picked up by a future W3C Working Group devoted to extend RDF. First of all, the evolution of RDF Schema to introduce constructs already present in other metadata definition languages. Related to this issue is the representation of measurement units in RDF to create units-aware descriptions of resources. Secondly, the introduction of new machine and human-readable syntaxes for RDF. Then, the interoperability with XML and JSON to foster the creation of Web Applications that exploit the Open Linked Data initiative. And, finally, the definition of a language and technology independent API for manipulating RDF, the "RDFOM".
Telefónica is one of the world’s largest telecommunications companies by market cap. Its activities are centred mainly on the fixed and mobile telephony businesses, while its broadband business is the key growth driver underpinning both. It operates in 25 countries and its customer base exceeds 264 million globally. Telefónica’s growth strategy is focused on the markets in which it has a strong foothold: Spain, Europe and Latin America.