Archive

Posts Tagged ‘open data’

Tell Me About Hampshire – Linking Government Data using SPARQL federation 2

March 23, 2014 3 comments

Yesterday I blogged about how to do some SPARQL federated queries across various government websites, and this blog is a continuation of this with a different example. In this blog I give an example query which basically say ‘tell me stuff about Hampshire‘. I do this by linking up data from Ordnance Survey, the Office of National Statistics, the Department of Communities and Local Government and Hampshire County Council. This query is really just for illustrative purposes, but I want to ask ‘for all districts in Hampshire find me the index of multiple deprivation rank, the change order and operative date for that district, the website for the local authority of that district along with the addresses of parcels of land where it is planned to build new dwellings. To achieve this I need to take data from several sources and use SPARQL federation. Here is the query that answers my question. First I query Ordnance Survey linked data to find districts in Hampshire, and I then pass these districts to three other linked data services to retrieve the relevant information. To try this example head over to the Ordnance Survey SPARQL endpoint and copy/paste the following:

select ?districtname ?imdrank ?changeorder ?opdate ?councilwebsite ?siteaddress
where
{?district <http://data.ordnancesurvey.co.uk/ontology/spatialrelations/within>
   <http://data.ordnancesurvey.co.uk/id/7000000000017765> .
  ?district a <http://data.ordnancesurvey.co.uk/ontology/admingeo/District> .
  ?district <http://www.w3.org/2000/01/rdf-schema#label> ?districtname .
 SERVICE <http://opendatacommunities.org/sparql> {
 ?s <http://purl.org/linked-data/sdmx/2009/dimension#refArea> ?district .
?s <http://opendatacommunities.org/def/IMD#IMD-rank> ?imdrank .
 ?authority <http://opendatacommunities.org/def/local-government/governs> ?district .
 ?authority <http://xmlns.com/foaf/0.1/page> ?councilwebsite .
 }
 ?district <http://www.w3.org/2002/07/owl#sameAs> ?onsdist .
 SERVICE <http://statistics.data.gov.uk/sparql> {
 ?onsdist <http://statistics.data.gov.uk/def/boundary-change/originatingChangeOrder>
          ?changeorder .
 ?onsdist <http://statistics.data.gov.uk/def/boundary-change/operativedate>
          ?opdate .
 }
 SERVICE <http://linkeddata.hants.gov.uk/sparql> {
   ?landsupsite <http://data.ordnancesurvey.co.uk/ontology/admingeo/district> ?district .
   ?landsupsite a <http://linkeddata.hants.gov.uk/def/land-supply/LandSupplySite> .
   ?landsupsite
<http://www.ordnancesurvey.co.uk/ontology/BuildingsAndPlaces/v1.1/BuildingsAndPlaces.owl#hasAddress>
   ?siteaddress .
   }
}

Happy SPARQLing…

Announcing new beta Ordnance Survey Linked Data Site

April 25, 2013 1 comment

Ordnance Survey has released a new beta linked data site. You can read the official press release here.

I thought I’d write a quick (unofficial) guide to some of the changes. The most obvious one that is hopefully apparent as you navigate round the site is the much improved look and feel of the site. Including maps (!) showing where particular resources are located. Try this and this for example. Maps can be viewed at different levels of zoom.

Another improvement is the addition of new APIs. The first of these is an improved search function. Supported fields for search and some examples can be found here. The search API now includes a spatial search element.

The SPARQL API is improved. Output is now available in additional formats (such as CSV) as well as the usual SPARQL-XML and SPARQL-JSON. Example SPARQL queries are also included to get users started.

Another interesting addition is a new reconciliation API. This allows developers to use the Ordnance Survey linked data with the Open Refine tool. This would allow a user to match a list of postcodes or place names in a spreadsheet to URIs in the Ordnance Survey linked data.

In the new release the Ordnance Survey linked data has been split into distinct datasets. You could use the above described APIs with the complete dataset or, if preferred, just work on the Code-Point Open or Boundary Line datasets.

For details on where to send feedback on the new site please see the official press release here.

Update: I blogged a bit more about some of the new APIs here.

Adventures with Kasabi…and a request for help…

October 5, 2011 1 comment

I’ve been playing around with Kasabi a bit of late. Kasabi is a new information market place from Talis that provides a useful place to publish your data, and then build services on top of it.

By way of a quick example you’ll see that the currently Ordnance Survey Linked Data is hosted in Kasabi here. There are a number of standard APIs provided with each dataset such as a SPARQL endpoint, search API etc. Kasabi provides an easy way to map complex SPARQL queries to simple API queries. For example, this API provides an easy way to do topological queries on the Ordnance Survey Linked Data. For example:

http://api.kasabi.com/dataset/ordnance-survey-linked-data/apis/33m?id=7000000000037256&spatialrelation=touches&apikey=yourkey

will find all regions that touch The City of Southampton.

This API gives you a list of all postcodes (and their lat/long) within a particular region. For example:

http://api.kasabi.com/api/ordnance-survey-postcode-region?district=7000000000037256&apikey=yourkey

will find all postcodes inThe City of Southampton. Additionally we can apply a style sheet to get back the same information as KML:

http://api.kasabi.com/api/ordnance-survey-postcode-region?district=7000000000037256&apikey=yourkey&output=kml

Kasabi also provides a handy place to store and host data, and one Sunday afternoon I decided to see how easy it would be to create a couple of hyperlocal datasets: one for Southampton and one for Hampshire. The basic approach in creating these hyperlocal datasets was to effectively ‘cookie cut’ region specific data from a number of different linked (open) data sets and put them into one store. So for Southampton and Hampshire we have a list of: airports, bus stops, stations, schools, GPs, hospitals, renewable energy generators, heritage sites, councillors, crime statistics, administrative regions and postcodes…   Each important element, like schools or hospitals, is linked to a postcode, district, ward—and in the case of Hampshire—county. The Ordnance Survey linked data is effectively acting as the clue between disparate sources of information. The fact that each of these datasets was provided as linked data, and furthermore referenced common identifiers for administrative regions and postcodes, meant it was very easy to bring them together in one store. Some sample queries are provided here. It gets more interesting when you combine elements from different datasets to, for example, ask questions like ‘find me GPs in my ward, and all the bus stops within a 100 metre radius of those GPs’.  If one were extra paranoid it would then be possible to extend the query to only find bus stops in areas of low anti-social crime levels. These queries are all well and good, or ‘nerdy, but nice’ as Andrew Stott put it.

What I was really hoping to do was build a nice webapp on top of this integrated data. Anyone who has seen my previous mash-ups will know that web design is not amongst my key skills, so Zach Beauvais suggested I put word out to the developer community to see if anyone fancied building on this data to make something cool and interesting like (for example) the (in)famous postcode paper. Any volunteers? :)

 

 

 

/location /location /location – exploring Ordnance Survey Linked Data – Part 2

October 25, 2010 5 comments

Ordnance Survey have now released an update to their linked data, which can be seen here. The new data now includes postcode information as well as a few changes to the administrative geography data. In this post I’ll go through what’s in the data, and give a few sample SPARQL queries.

I spoke a bit about the administrative geography data in a previous blog post – but the data has changed a bit since then. Just to re-cap the administrative geography linked data contains information about administrative and voting geographic regions. These include unitary authorities, counties, wards, constituencies, Welsh Assembly regions and a whole lot more [1]. Here are some examples:

If you want to find a full list of the sorts of thing you can find in the data simply go to the query interface (or SPARQL endpoint as it is know) and try the following query:

select distinct ?type

where { ?a a ?type . }

Now you have the list all of type of things in the data you can as for lists of instances of those types.

For example, the following query will return all of the unitary authorities:

select ?a

where {

?a a <http://data.ordnancesurvey.co.uk/ontology/admingeo/UnitaryAuthority&gt; .

}

All of the names of all the regions have now been modelled using the SKOS vocabulary. If you want to find the official names of all the unitary authorities you can simple issue a query like:

select ?a ?name

where

{

?a a <http://data.ordnancesurvey.co.uk/ontology/admingeo/UnitaryAuthority> .

?a <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  .}

Also included in the data are two attributes called Unit ID and Area Code. These values are useful if you want to produce a mashup using this data and display it by boundary.

So for example, for Southampton (http://data.ordnancesurvey.co.uk/id/7000000000037256) the area code is UTA (for unitary authority) and the unit ID is 37256. These values can be used as follows:

/*here we set-up the our variable called ‘boundaryLayer’ with the strategies that we require. In this case, it is its ID and type i.e. Unitary Authority */

boundaryLayer = new OpenSpace.Layer.Boundary(“Boundaries”,

{ strategies: [new OpenSpace.Strategy.BBOX()], admin_unit_ids: ["37256"], area_code: ["UTA"] });

//then we add the bounadry to the map osMap.addLayer(boundaryLayer);

//this effectively refreshes the map, so that the boundary is visible

osMap.setCenter(osMap.getCenter());

to display the Southampton boundary using the OS OpenSpace API. See http://openspace.ordnancesurvey.co.uk/openspace/support.html for more details.

Arguably the most useful information in this data are the qualitative spatial relationships between different regions. Regions are related to the regions they contain, they are within and they touch. In the case of the touching relationship only regions of the same type have an explicit touching relationship. The exception to this are unitary authorities, counties, district and metropolitan district that also have touching relationships between each other. The following simple query will return a list of all counties, districts and unitary authorities that border The City of Southampton. It will also return their names:

PREFIX spatialrelations: <http://data.ordnancesurvey.co.uk/ontology/spatialrelations/&gt;

select ?a ?name

where

{

?a spatialrelations:touches <http://data.ordnancesurvey.co.uk/id/7000000000037256&gt; .

?a <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  .

}

If you are only interested in the bordering counties you can add an extra line to your query:

PREFIX spatialrelations: <http://data.ordnancesurvey.co.uk/ontology/spatialrelations/&gt;

select ?a ?name

where

{

?a spatialrelations:touches <http://data.ordnancesurvey.co.uk/id/7000000000037256&gt; .

?a <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  .

?a a <http://data.ordnancesurvey.co.uk/ontology/admingeo/County> .

}

Similarly, the following query returns all the county electoral divisions (and their names) within Hampshire:

PREFIX spatialrelations: <http://data.ordnancesurvey.co.uk/ontology/spatialrelations/&gt;

select ?a ?name

where

{

?a spatialrelations:within <http://data.ordnancesurvey.co.uk/id/7000000000017765&gt; .

?a <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  .

?a a <http://data.ordnancesurvey.co.uk/ontology/admingeo/CountyElectoralDivision> .

}

For convenience some shortcuts have been added to the data in this release. For certain nesting geographies, such as the county – district – parish or district – ward nestings, various new properties have been added. For example, the property ‘counyElectoralDivision‘ relates all counties to their constituent county electoral divisions. The above query can now be done in a simpler way:

PREFIX admingeo: <http://data.ordnancesurvey.co.uk/ontology/admingeo/&gt;

select ?a ?name

where

{

<http://data.ordnancesurvey.co.uk/id/7000000000017765&gt; admingeo:countyElectoralDivision ?a .

?a <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  .

}

Similar predicates such as ‘county‘, ‘district‘, ‘ward‘, ‘constituency‘ etc. provide similar shortcuts. For example, the following returns all the Westminster constituencies in South East England.

PREFIX admingeo: <http://data.ordnancesurvey.co.uk/ontology/admingeo/&gt;

select ?a ?name

where {

<http://data.ordnancesurvey.co.uk/id/7000000000041421&gt; admingeo:westminsterConstituency ?a .

?a <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  . }

The most significant introduction to this data is the inclusion of postcode information. The data now contains information about postcode units, postcode sectors, postcode districts and postcode areas. For each postcode unit an easting/northing coordinate value is given [2] along with the district, ward and county (where applicable) that contains said postcode unit. An example of this can be seen for the Ordnance Survey postcode SO16 4GU. Each postcode is also related to its containinb postcode area, sector and district.

The properties ‘ward‘, ‘district‘ and ‘county‘ relate a postcode to the relevant regions. The simple query:

PREFIX postcode: <http://data.ordnancesurvey.co.uk/ontology/postcode/&gt;

select ?district

where {

<http://data.ordnancesurvey.co.uk/id/postcodeunit/SO164GU&gt; postcode:district ?district .

}

returns the unitary authority that contains the postcode SO16 4GU.

This query:

PREFIX spatialrelations: <http://data.ordnancesurvey.co.uk/ontology/spatialrelations/&gt;

select ?postcode

where

{

?postcode spatialrelations:within <http://data.ordnancesurvey.co.uk/id/postcodearea/SO&gt; .

}

returns all the postcodes in the SO postcode area.

We can combine the above two queries to find the areas, along with their names, covered by the postcode area SO:

PREFIX spatialrelations: <http://data.ordnancesurvey.co.uk/ontology/spatialrelations/&gt;

PREFIX postcode: <http://data.ordnancesurvey.co.uk/ontology/postcode/&gt;

select distinct ?district ?name

where

{

?postcode spatialrelations:within <http://data.ordnancesurvey.co.uk/id/postcodearea/SO&gt; .

?postcode postcode:district ?district .

?district <http://www.w3.org/2004/02/skos/core#prefLabel&gt; ?name  .

}

Hopefully these few examples will give you enough information to fully explore this new release of the Ordnance Survey linked data. For those of you who don’t like SPARQL watch this space – hopefully we will soon(ish) have an API built on top of this data to allow for even easy access.

[1] you’ll notice the ‘isDefinedBy’ link currently returns a 404 – not for long I hope :)

[2] lat/long to follow

Follow

Get every new post delivered to your Inbox.

Join 2,189 other followers