Where do I find a list of of UK public libraries so that I can:
a) plot them on a map;
b) search for them by location (e.g. within five miles of a particular postcode)?
For UK Onine Centres, it’s easy enough (5 Minute Hack – UK Centres Online Map); and there’s the start of a hack for plotting out the locations of all HEI libraries via SCONUL data (UK HE LIbraries map); but when I asked the lazyweb this question with respect to public libraries at the end of last week, the best response I got back was to use Librarything (Libraries Near Me Map (Courtesy of LibraryThing)).
So how about data.gov.uk? Has any library data been logged there? Searching for library didn’t turn up much at all, though libraries was a little more successful; here’s the result for library OR libraries
More specifically, at the current time, there are the following relevant datasets:
A list of libraries in Warwickshire, including contact details Overview … Tags warwickshire-county-council libraries warwickshire Department … Council Wiki Libraries Resources Other XML …
List of all Libraries in Sunderland. Overview … Tags local-authority-services libraries sunderland Department … Wiki Sunderland Libraries Resources …
East Staffordshire Libraries
A list of the libraries in East Staffordshire. Overview … – Tags libraries east-staffordshire Department … Wiki East Staffordshire Libraries Resources Other XML …
Details of Bournemouth Libraries, including location, opening times Overview … Tags bournemouth-borough-council libraries local-authority bournemouth … Wiki Bournemouth Libraries Resources …
Hampshire County Council Recreation Sites
… gardens reserve arts country-park library nature Department … Hampshire County Council, hampshire-county-council, library, museum, nature, reserve) …
The data is provided in a range of formats (I think I really need to do a set of recipes for handling each type..?):
– Sunderland libraries use another recognised XML format, the ESRI ArcGIS 9.2 XML representation:
– Warwickshire libraries look to be using their own brand of XML, giving name, structured address, phone, fax, email, web address, co-ordinates (a lat, long pair) and a photo:
– Bournemouth Libraries publish their data as CSV including name, address, phone, email, an empty A column (A HREF=…? ie a slot for URLs?), and opening times by day of week; the geodata is provided as eastings and northings:
– Hampshire County Council list their libraries in the context of recreational sites (cultural and leisure services is another description that often gets used, I think?) as a CSV file, with headers Name,Address,Town,Phone,Postcode,,Lat/long,URL,Short description,Tags. (NB there is an unlabeled column in there – the lat/long column header should be split over two columns – one for lat, one for long); libraries are identified from the description column (Local Library) and from the tag column (library).
So – what can we say? Five data sets describing the location of libraries from five councils using five different representations. Which is best? Which is easiest to generate? Which is easier to use, for what, by whom?
I think I need a couple more posts on this topic – one looking at ways of actually getting a quick map based preview of the data in the data sets, one looking at the issues involved with generating and publishing the data on the council side. (If any council developers read this, I’d appreciate your comments about how you would go about generating these data files and what issues there are wrt getting the data out in different formats.)
PS One of the many map related things I haven’t tried yet (the following was one of the things I wanted to do had I made it to today’s Chips and Mash Mashed Library event) is to have a play the Google (yes, I know… sorry…) Maps data api, which allows you to import a geodata set as KML or CSV data, and then use the API “to filter a large set of features by radius, bounding box, or text attributes, and sort them by their distance from a location”. I’m not sure what the licensing is around the scraped SCONUL library location data, but it’d be a good candidate for a practical, proof-of-concept play with the Google Maps data API… It would also be interesting to see how well it copes with the two flavours of CSV (one using lat/long, one with eastings/northings) described above…