Technologies of linked data: URIs

The following post is an excerpt from my thesis entitled Linked open data for public sector information.
Uniform Resource Identifiers (URI) offer an extensible, federated naming system for universal and global identification [1, p. 6]. Thanks to URI’s universality, resource identified with a URI may be anything, including web sites, ideas, and real-world objects.
URIs and Uniform Resource Locators (URLs) are different. URI needs not to locate the resource it identifies. Location of a resource is described by a URL, that in addition to identifying the resource provides a way of addressing it. In some cases, a resource may have the same URL as URI. This is true for information resources that may be retrieved via the Web. However, resources that may not be retrieved via the Web, such as physical objects, have a URI but do not have any URL, since they cannot be located in that way.
Resource needs not to be identified with a single URI because linked data adopts the non-unique name assumption allowing equivalent resources to have multiple URIs. This approach lowers the start-up barriers for data modelling since it lets linked data publishers to assign resources with their own URIs instead of making the effort to find the URIs that already exist for such resources.

References

1. RFC 3986. Uniform Resource Identifier (URI): generic syntax [online]. BERNERS-LEE, Tim; FIELDING, Roy Thomas; MASINTER, Larry. January 2005 [cit. 2012-04-23]. 61 p. Available from WWW: http://tools.ietf.org/html/rfc3986. ISSN 2070-1721.