Implementation of Identifiers using URIs in INSPIRE – Frequently Asked Questions

Why does INSPIRE recommends to use URIs in the http scheme?

Originally, INSPIRE was not restricted to the web and other networks and identification mechanisms were considered as well. Looking at the implementation, however, both INSPIRE and the web have moved on. In INSPIRE, there is no longer an expectation that there will be any other network than the web. In the web, http URIs have become the primary way to reference information resources on the web. To keep INSPIRE connected with the development of related information infrastructures, in particular in the e-government context, the use of http URIs is considered prudent.

http URIs do not require that the URI can be dereferenced, although it is good practice to support dereferencing for information resources like spatial objects.

What are the benefits of URIs?

The URI is relatively easy to deploy and allows to build INSPIRE on the powerful and highly scalable infrastructure of the web. Integrating INSPIRE in the web architecture supports also the reuse of spatial data from INSPIRE by other communities as familiar technologies and protocols are used.

Proper URI design is important to support the sustainability of the URIs. See the URI guide in D2.5 Annex H.          

How are the different identifiers encoded in the data?

In the GML representation:

• the URI's are encoded in the gml:identifier property of each spatial object or data set
• unique object identifiers are encoded as inspireId attributes of the spatial object and use a data type base:Identifier (note: since the inspireId information is also encoded in the gml:identifier property it is currently under discussion to drop the inspireId property from the GML encoding)
• thematic identifiers (identifiers of the real-world phenomenon) are encoded as standard attributes of the spatial object, with the Annex II/III data models a new data type base2:ThematicIdentifiers has been introduced for such attributes

What is the role of URIs in download services?

URIs are used to identify and reference spatial data sets. Dereferencing these URI's will result in downloading files representing the spatial data sets. There may be multiple URIs for a data set, e.g. to provide the same data set in multiple coordinate reference systems or languages.

Besides downloading the data sets, INSPIRE download services are recommended to also provide access to individual spatial objects in which case dereferencing a URI of a spatial object will return a representation of the spatial object. The implementation details will be transparent to user / client, but may, for example, be implemented by a redirect to a GetFeatureById query to an INSPIRE download service based on the OGC Web Feature Service (WFS) specification. I.e., the URIs are thus well suited to using data “in situ”, too.

What other types are identified using URIs beside spatial objects?

A number of shared resources are identified as part of INSPIRE and are referenced from data sets in INSPIRE. These resources are usually maintained in a register and are identified by URIs, for example:

• Code list values, e.g. http://inspire.ec.europa.eu/codelist/countrycode/de
• Coordinate reference systems, e.g. http://www.opengis.net/def/crs/EPSG/0/4258

Are there examples of INSPIRE URIs I can look at?

Yes several Member States are developing data sets using URI, for example:

• the United Kingdom: see Designing URI Sets for Location

Who allocates URIs?

Once a domain has been registered at national level (or by an international organisation), ideally by the appropriate coordinating body and namespaces agreed, the data provider is then able to continue to allocate URI and publish their data.

Is INSPIRE going to organise a URI framework?

No, the framework exists since INSPIRE is reusing the existing web resources. Ideally each member state (and perhaps the EU for any truly pan-European datasets), will establish their own domain and manage their own resources. This will reduce costs and ensure member states are not limited by any third party resources.

We do not have any URI mechanisms defined at national level in my member state – how do I get started?

Approach your national coordinating body in the first instance. Also look at what other Member States have been doing and use opportunities to discuss this in pan-European INSPIRE groups.

Can I use thematic identifiers for spatial objects?

Thematic identifiers are used as attributes on spatial objects and should be published where available, but since they describe real world phenomena they are unsuited for use as a spatial object identifier.

Do URIs carry any metadata?

It is possible to include some metadata in the URI or as a link to the URI (which can dereferenced from the URI).

Can I version URIs?

Yes, a versioning mechanism can be added to the URI and is recommended to include the rules in the URI scheme.

Can I change the URI structure a few years’ time if I need to??

While it is possible to do this – persistence of identifiers is critical for the infrastructure and the development of future information services. Users will build their services around these identifiers and expect that they continue to work.

URIs can be “redirected” when dereferenced and one way to decouple the URI scheme from the implementation details is to use this standard capability of HTTP. However, as redirection has an impact on performance, redirection should be used with care.

As any significant change will incur costs to data providers and users, it is recommended to design a simple approach and maintain that going forward, where possible.

Will using URIs mean I cannot protect my data from unauthorised download?

You may still protect your data using standard web security mechanisms – though that will clearly limit the wider benefits of publication in support of reuse.

What do I get back when I follow a URI?

This depends on what and how the data publisher has configured their data. It is possible to publish data in several formats e.g. HTML, GML, JSON, RDF, Shapefiles, GeoTIFF, etc. HTTP supports content negotiation between the client and the server. Most data providers will probably start with satisfying the minimum requirement and extend this over time based on demand.

How will users use my data with URI?

In all sorts of ways – often using Web APIs and applications since the URI are primarily “machine readable”. This reinforces the need to maintain a consistent and simple approach in publishing data.

The INSPIRE discovery service is designed to locate datasets – how will a user find a discrete spatial object using a URI?

If a user already has a URI of a spatial object, he can continue to use it to reference the object also in other contexts.

If a data set is provided by a download service that supports direct access, a specific spatial object of interest can be identified using queries, e.g. using the location, thematic identifiers or other properties. Once identified, the URI can be used to reference the spatial object.

For other data sets, the whole data set may have to be downloaded and inspected to identify the URI of a spatial object of interest.