URI Minting

In Linked Data URIs and their internationalized generalization IRIs are used to identify things, i.e., "resources", in the WWW and outside it. IRIs are an extension of URIs that permits the use of characters outside those of plain ASCII. Although there are three non-ASCII letter commonly used in Finnish and Swedish (ä, ö, and å), we prefer sticking to using only ASCII charters in identifiers and hence use URIs. URIs are, for example, easy to input from different keyboards and avoiding non-ASCII characters avoids problems in many programming environments.

One of the key ideas of Linked Data is that an URI not only unambiguously and globally identifies a resource (using the name server infrastructure of the WWW), but is also a Web address by which more data about the resource can be found by 1) a computer and 2) by a human data user. The HTTP protocol of the WWW is used for this. Best practices for URI data services are discussed, e.g., in the W3C Note Cool URIs for the Semantic Web.

URI Minting Policy in LDF

Our policy for minting IRIs is to use the template below (i.e., we use "303 IRIs", not "hash IRIs"; cf. discussion in the W3C Note above):

http://ldf.fi/DATASET/ID

Here DATASET is the name of a dataset, e.g., "my-data". ID a is local name starting with a letter and ending with an integer, e.g., "p3765". An example of an URI is:

http://ldf.fi/my-data/p3765

Numeric local names without "meaning" are used because it is usually considered a good practice in LD not to embed semantics in local names, because this creates less pressure to change IRIs if semantics change in the underlying world (URIs should be as persistent as possible). However, for schemas, mnemonic names (preferably in English) can are used, as customary in, e.g., Semantic Web standards, such as RDF, RDFS, and OWL.