Web resource: Difference between revisions

Revision as of 05:49, 2 February 2007

Revision as of 20:25, 16 February 2007

Line 25:

Building on this recursive principle, RDF vocabularies, such as [[RDFS]], [[Web Ontology Language |OWL]], and [[SKOS]] will pile up definitions of abstract resources such as classes, properties, concepts, all identified by URIs.

RDF also specifies the definition of anonymous resources or [[Blank node | blank nodes]], which are not absolutely identified by URIs.

=== Using HTTP URIs to identify abstract resources ===

Using URLs, and singularly HTTP URIs, to identify abstract resources, such as classes, properties or other kind of concepts, is a frequent practice, for example in RDFS or OWL [[Ontology (computer science) |ontologies]]. Since such URIs are associated with the HTTP protocol, the question arose of which kind of representation, if any, should be get for such resources through this protocol, typically using a Web browser, and if the syntax of the URI itself could help to differentiate "abstract" resources from "information" resources. The URI specifications such as RFC 3986 let to the protocol specification the task of defining actions performed on the resources and they don't provide any answer to this question. It had been suggested that http URIs identifying a resource in the original sense, file, document or any kind of so-called information resource, should be "slash" URIs, in other words should not contain [[fragment identifier |fragment identifiers]], whereas URIs used to identify concepts or abstract resources should be "hash" URIs using fragment identifiers. For example <nowiki>''http://www.sillywidgets.org/catalogue/widgets.html''</nowiki> would both identify and locate a web page (maybe providing some human-readable description of the widgets sold by Silly Widgets, Inc.) whereas <nowiki>''http://www.widgets.org/ontology#Widget''</nowiki> would identify the abstract concept or class "Widget" in this company ontology, and would not necessarily retrieve any physical resource through http protocol. But it has been answered that such a distinction is impossible to enforce in practice, and famous standard vocabularies provide counter-examples widely used. For example the [[Dublin Core]] concepts such as "title", "publisher", "creator" are identified by "slash" URIs like ''http://purl.org/dc/elements/1.1/title''.

The general question of which kind of resources http URI should or should not identify has been formallyformerly known in W3C as the [http://www.w3.org/2001/tag/issues.html#httpRange-14 httpRange-14] issue, following its name on the list defined by the [[Technical Architecture Group]] (TAG). The TAG has delivered in 2005 a final answer to this issue, making the distinction between an "information resource" and a "non-information" resource dependent on the type of answer given by the server to a "GET" request. This solution puts and end to the "hash" vs "slash" debate, and seems to have met a consensus in the Semantic Web community, although some of its prominent members such as [[Patrick J. Hayes | Pat Hayes]] have expressed concerns both on its technical feasibility and conceptual foundation. According to Patrick Hayes' view point, the very distinction between "information resource" and "other resource" is impossible to found, and should better not be specified at all, and [[ambiguity]] of the [[referent]] resource is inherent to URIs like to any naming mechanism.

=== Resource ownership, intellectual property and trust ===