Linked Open Data & Metadata

Abstract

This article considers linked data, starting with the four rules drawn up in 2006 by the inventor of the web, Tim Berners-Lee, to produce this kind of data: (1) to use a web standard, the Internationalized Resource Identifier (IRI), to name things within the data; and, in particular, (2) to use IRIs of the HTTP protocol, so that data associated with these IRIs can be retrieved and accessed in exactly the same way that web pages are retrieved and accessed; (3) to use a second web standard, the Resource Description Framework (RDF), to format the data, and a third web standard, the SPARQL Query Language, to query those data. Finally, (4) to use IRIs from other datasets within the data, so as to connect one’s own data with those of other people. This article comments on these rules and discusses their implications, highlighting in particular the fact that they lay the bases for the creation of the Semantic Web, that is, a new web, parallel to the web for humans as we have known it for the last two decades. The Semantic Web is made up of pages containing formal knowledge expressed as linked data. This knowledge is consumed by artificial agents carrying out trivial, time-consuming, and error-prone tasks (such as counting the occurrences of a certain syntactic construct in Dante’s Comedia), freeing humans from such tasks and letting them use their time for more intellectual activities (such as figuring out the evolution of Dante’s culture). The vision of the Semantic Web is presented along with two basic ingredients for its establishment: the Resource Description Framework (RDF), the language for expressing linked data, and ontologies – that is, vocabularies – that axiomatize the definitions of the terms used in linked data. For the realization of the Semantic Web, RDF is necessary but not sufficient, because RDF provides the mere structure of linked data, without indicating any particular way to represent a specific domain. This is the role of ontologies, without which any linked data dataset would remain confined within a (possibly very small) community, defeating the vision of a common, global data space. Finally, the article discusses the role of the Semantic Web for the scholarly domain. In fact, linked open data and ontologies play a very important role in the scientific and scholarly world by offering tools for the creation and sharing of data and vocabularies. The key concept here is interdisciplinarity. It has been long recognized that significant progress can be achieved in all branches of science in research projects that are able to combine tools, data, and knowledge from different domains. Research infrastructures such as D4Science are complex systems that allow users to realize interdisciplinarity in science by offering scientists virtual research environments where they can find the tools, data, and knowledge that they need for their work. They also provide them with the communication and collaboration facilities that are necessary to cooperate with their colleagues. D4Sceince is also supporting the humanities with virtual research environments like those of the PARTHENOS and ARIADNE infrastructural projects.