This is a basic RDF vocabulary that provides the Semantic Web community with a namespace for representing lat(itude), long(itude) and other information about spatially-located things, using WGS84 as a reference datum.
The vocabulary is getting significant usage, both (as intended) within RDF documents, but also as a namespace used within non-RDF XML documents, such as RSS 2.0 (see below).
The Locative packets format uses this vocabulary, as do Map Bureau's RDF mapping tools. The geocoder.us site, provides a free geocoding service for the US, based on TIGER data, and accessible via an RDF Web service that uses this vocabulary. Other applications include blogmapper and openguides. The WordKit system also uses it for geocoding RSS 1.0 (ie. RDF), RSS 2.0 (non-RDF) and Atom (non-RDF). The Yahoo! Maps service also makes use of this namespace, although apparently only within a RSS 2.0 context. See the ESW:GeoInfo Wiki entry for links to other uses of this work.
This document was created as an informal collaboration within W3C's Semantic Web Interest Group. This work is not currently on the W3C recommendation track for standardization, and has not been subject to the associated review process, quality assurance, etc. If there is interest amongst the W3C membership in standards work on a location/mapping RDF vocabulary, this current work may inform any more formal efforts to follow.
This vocabulary begins an exploration of the possibilities of representing mapping/location data in RDF, and does not attempt to address many of the issues covered in the professional GIS world, notably by the Open Geospatial Consortium (OGC). Instead, we provide just a few basic terms that can be used in RDF (eg. RSS 1.0 or FOAF documents) when there is a need to describe latitudes and longitudes. The motivation for using RDF as a carrier for lat/long info is RDF's capability for cross-domain data mixing. We can describe not only maps, but the entities that are positioned on the map. And we can use any relevant RDF vocabularies to do so, without the need for expensive pre-coordination, or for changes to a centrally maintained schema.
A basic, standalone example:
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#"> <geo:Point> <geo:lat>55.701</geo:lat> <geo:long>12.552</geo:long> </geo:Point> </rdf:RDF>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns="http://xmlns.com/foaf/0.1/"> <Person> <name>Dan Brickley</name> <homepage dc:title="Dan's home page" rdf:resource="http://danbri.org/"/> <based_near geo:lat="51.47026" geo:long="-2.59466"/> <rdfs:seeAlso rdf:resource="http:/danbri.org/foaf.rdf"/> <!-- more RDF here, using any RDF vocabularies --> </Person> </rdf:RDF>
An example (from WorldKit) of geo-coding with RSS 1.0:
The origin of this workspace was the 2003-01-09 discussion in the RDF Interest Group IRC channel. We are taking a similar approach to the exploration of RDF vocabulary for calendars, although that work is more mature. See the RDF Calendar Workspace for more information.
Currently we specify only a very minimalistic RDF vocabulary for describing Points with latitude, longitude, and altitude properties from the WGS84 reference datum specification.
This design allows for basic information about points to be described in RDF/XML, and augmented with more sophisticated or application-specific metadata.
The vocabulary also defines a property lat_long, but this should probably be removed, as few commentators have valued it.
The vocabulary defines a class 'Point', whose members are points. Points can be described using the 'lat', 'long' and 'alt' properties, as well as with other RDF properties defined elsewhere. For example, we might use an externally defined property such as 'bornNear' or 'withinFiveMilesFrom', or perhaps other properties for representing lat/long/alt in non-WGS84 systems.
The 'lat' and 'long' properties take literal (ie. textual values), each in decimal degrees. The 'alt' property is decimal metres about local reference ellipsoid.
Whitespace in property values is insignificant, and
discouraged. We do not use RDF's datatyping mechanism in the
vocabulary's schema to note that the
properties are XML Schema
float datatypes. Instead,
from RDF's point of view, the properties are simply strings.
This reduces the syntactic burden on RDF/XML documents using
our vocabulary, since we write lat and long information as
strings. The string representation of lat and long should follow
the rules for XML Schema
float, even though we do
not indicate explicitly in RDF that we are representing floating
point numbers. A future version of this document may be stricter
about whitespace when marking up
float values, so it
is probably best to avoid any whitespace in your lat/long/alt
The semantics of the lat/long/alt properties are unaffected by xml:lang values, if present.
While it is quite possible to use the namespace in non-RDF XML documents, developers should note that this reduces the availability of such data within the Semantic Web, since specialist knowledge of each XML tagset is needed before data can be safely interpreted. The GRDDL specification offers an XSLT-based mechanism for mapping such data into RDF.
As this is an RDF vocabulary, we define properties of a kind of thing, a Point, rather than the structure and characteristics of an XML document type. XML Schemas and DTDs typically adopt the latter approach, and thus provide more guarantees about the information content of each document of some type. By contrast, our RDF vocabulary provides no guarantee about which things will be described in any particular RDF/XML document. For example, it may turn out that many point descriptions omit the altitude property, 'alt'. This doesn't make those documents invalid in any sense; they are merely less informative.
We can note some rules for reasoning about the identity conditions for points. If we encounter a description of a Point, call it ?X, and another, call it ?Y, and ?X and ?Y have identical values for their 'lat' and 'long' and 'alt' properties, we can conclude that ?X == ?Y, ie. that ?X and ?Y represent the self-same thing. Anything that has any 'lat', 'long' or 'alt' properties will be a 'Point'. In our RDF Schema for the vocabulary, we indicate this using the rdfs:domain property.
A Point has only one 'lat', only one 'long', and only one 'alt'; we could use W3C's Web Ontology Language (OWL) to express this. OWL allows us to note that these RDF properties are 'functional properties'.
Regarding the http://www.w3.org/2003/01/geo/wgs84_pos# vocabulary: it is pretty simple and should hopefully stay that way, ie. short and simple enough to memorise and use. There may be some need to extend it, based on deployment experience, test cases etc. Time will tell.
$Log: Overview.html,v $ Revision 1.21 2006/02/01 22:14:46 danbri added a parag noting decimal degrees for lat and long; decimal metres above ellipsoid for alt Revision 1.20 2005/07/08 18:22:34 danbri A few more introductory words. Revision 1.19 2005/07/08 18:14:35 danbri added links to rss1, dublincore, foaf pages Revision 1.18 2005/07/08 18:01:08 danbri s/rdfig/swig/ somewhere i missed Revision 1.17 2005/07/08 17:59:11 danbri trimmed Revision 1.16 2005/07/08 17:55:54 danbri more links Revision 1.15 2005/07/08 17:45:30 danbri added geocoder link Revision 1.14 2005/07/08 17:44:41 danbri added geocoder link Revision 1.13 2005/07/08 17:19:35 danbri markup bug Revision 1.12 2005/07/08 17:18:25 danbri noted that 2nd eg uses DC too Revision 1.11 2005/07/08 17:17:08 danbri CSS Revision 1.10 2005/07/08 17:14:35 danbri CSS Revision 1.9 2005/07/08 17:11:56 danbri CSS Revision 1.8 2005/07/08 17:11:06 danbri CSS Revision 1.7 2005/07/08 17:05:26 danbri more examples Revision 1.6 2005/07/08 16:55:36 danbri tidying text Revision 1.5 2005/07/08 16:54:25 danbri moved nonrdf later into doc Revision 1.4 2005/07/08 16:51:59 danbri markup fixes Revision 1.3 2005/07/08 16:50:41 danbri more text, discussion of non-rdf xml, and links to apps Revision 1.2 2005/07/08 16:33:23 danbri added more navigation Revision 1.1 2005/07/08 16:25:23 danbri new writeup Revision 1.38 2004/02/06 17:38:12 danbri Fixed a bad commit screwup Revision 1.36 2003/12/14 02:23:14 danbri anchored Revision 1.35 2003/12/14 02:19:29 danbri anchored Revision 1.34 2003/04/03 00:30:01 danbri removed related links; linked to Wiki instead Revision 1.33 2003/02/19 00:40:10 danbri removed SuckyCaps from Grub[Ss]treet (even if it is WikiBased :) Revision 1.32 2003/01/11 19:02:23 danbri Updated datatypes section, now that datatyping has been removed from the schema. Revision 1.31 2003/01/11 12:27:52 danbri Added link to 2.2M Danish addresses. Revision 1.30 2003/01/11 12:09:12 danbri Added Sean Palmer's work on computing distances in RDF, and his infomesh-geo schema. Revision 1.29 2003/01/11 12:03:33 danbri Added link to Sean Palmer's 'geo' URN proposal. Revision 1.28 2003/01/11 11:19:58 danbri Revised the note re whitespace to mention use of RDF datatyping, and to discourage use of whitespace. This should be re-organised to separate the two issues. Also noted need for clearer advice on the whitespace issue. Revision 1.27 2003/01/11 10:57:40 danbri Added final '#' to namespace uriref in one place that I missed yesterday. Revision 1.26 2003/01/11 01:23:25 danbri Added DAML Geofile data. Revision 1.25 2003/01/10 23:05:49 danbri Changed to say: *An* RDF Geo vocab... (there will doubtless be others). Revision 1.24 2003/01/10 22:52:40 danbri Fixed a bug in the identity reasoning section ('alt' was ommitted from the identity rules). Added a paragraph contrasting RDF and XML schema systems. Revision 1.23 2003/01/10 22:33:54 danbri Added example snippet. Also Syndic8 Geo metadata maker. Revision 1.22 2003/01/10 21:31:43 danbri Swapped a couple of paragraphs around, fixed some markup bugs and added a comment explaining the role of rdfs:domain. Revision 1.21 2003/01/10 21:28:29 danbri Re-organised the page, and added some more detailed description of the Point/lat/long/alt design, mentioning identity rules for the Point class and the possible use of OWL to clarify the design. Revision 1.20 2003/01/10 19:53:43 danbri Added DCMI Point Encoding Scheme. Revision 1.19 2003/01/10 19:46:54 danbri Adding final '#' to geo vocab namespace uriref, for consistency with schema. Revision 1.18 2003/01/10 19:45:28 danbri Added link to Syndic8 metadata spec. Revision 1.17 2003/01/10 17:58:22 danbri Added a link to GML 2.0, Simon Cox et al. for OGC., Feb 2001. Revision 1.16 2003/01/10 17:50:22 danbri Added Muxway links, from Joshua Schachter. Revision 1.15 2003/01/10 17:45:20 danbri Added @semantics link from Alberto. Revision 1.14 2003/01/10 16:51:46 danbri Added streetmap Perl module, and reference to xmethods.net web service directory. Revision 1.13 2003/01/10 16:43:03 danbri Added GeoURL location/add yourself page, NVML and DCMI Box Encoding Scheme. Revision 1.12 2003/01/10 16:09:45 danbri Added Spacenamespace sample data from Jo Walsh, and ptr to PostGIS docs. Revision 1.11 2003/01/10 14:15:14 danbri Noted that OSGB FAQ includes pointers to conversion code. Revision 1.10 2003/01/10 14:13:05 danbri Added Garmin, UK waypoints and OS links. Revision 1.9 2003/01/10 12:50:21 danbri added previous cvs history, so you can see how busy I've been... revision 1.8 date: 2003/01/10 12:48:14; author: danbri; state: Exp; lines: +8 -1 added blogmapper and rdfmap links revision 1.7 date: 2003/01/10 12:42:05; author: danbri; state: Exp; lines: +4 -1 added rdfmap revision 1.6 date: 2003/01/10 12:37:16; author: danbri; state: Exp; lines: +27 -8 added Status info revision 1.5 date: 2003/01/10 12:30:23; author: danbri; state: Exp; lines: +3 -1 added opengis consortium revision 1.4 date: 2003/01/10 11:59:03; author: danbri; state: Exp; lines: +3 -2 fixed buglet in vocab, added more wgs links revision 1.3 date: 2003/01/10 11:46:43; author: danbri; state: Exp; lines: +4 -4 added more links revision 1.2 date: 2003/01/10 11:40:50; author: danbri; state: Exp; lines: +5 -5 added more links revision 1.1 date: 2003/01/10 11:28:17; author: danbri; state: Exp; first cut at contextual overview doc