| 
  • If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

UK Cultural Heritage

Page history last edited by Mia 12 years, 3 months ago

Team

Mia Ridge

Keith May

Ceri Binding

Kate Byrne

Alex Dutton

 

Peer review (of sorts) of data sources from POV of someone who wants to find biggest overlapping sets of data about UK Cultural Heritage.

 

Case Study

We tried being someone looking for all info they can possibly find about a particular place: Greenwich. Summary of result: it wasn't easy to connect the datasets.

 

 

 

Starting from the viewpoint that a humanist might want to look over the data before working out the tech acronyms.

 

Discovering that everyone is using different ontologies so it's hard to match them up...

 

Some Resources to Consider

BM SPARQL endpoint http://collection.britishmuseum.org/Sparql/. 260 predicates and about 42 classes in schema.

ADS LoD resources http://data.archaeologydataservice.ac.uk/page/ plus their non-RDF stuff (grey lit)

Science Museum CSV files http://api.sciencemuseum.org.uk/documentation/collections/ 

National Maritime Museum http://collections.nmm.ac.uk/solr 

LAARC http://www.museumoflondon.org.uk/laarc/catalogue 

 

Issues we're noting as we go along

  • It's fairly easy to get some handle on the BM data because you can do what we did - query to find what the classes and predicates are, which gives you a rough idea of the structure. We had problems trying to get to grips with the ADS data because we couldn't do this.
  • problem with BM data for using PLACE to link datasets: the places seem to be locations within the museum... are we misunderstanding...?
  • Science Museum data has place of manufacture - we could try linking on this
  • CSV data is really nice and clear! :) 
  • shouldn't have to sacrifice user readability (browseability) for machine readability, the way we're finding we do - need to browse to see what might link, suss out formats, coverage, etc.

 

Starting with the task of finding out what's available in different datasets for 'Greenwich', then thinking about how we can parse them and display them in a shared interface... Sitecodes are possibly too specific to archaeology; grid references are UK-only and not generally supported by geo-referencing tools...

 

Can we link results from various resources into one interface?  Tried the following:

 

But for all these I've been reduced to using string searches rather than a place ID, grid reference, lat long, polygons, etc.

 

Final Session

 

So, we're now trying to do something practical - maybe looking for Greenwich-related connections between British Museum and Science Museum objects. We've been joined by Dominic Oldman of BM, which is helping us a lot!

 

For starters, Dominic has pointed us to useful documentation about their RDF data, at http://www.researchspace.org/Stage-2-Outputs. They're planning to put the schema up as well, showing the hierarchy and which bits of CRM they're using.

 

Tasks we have to do

  1. Query BM thesaurus to find how "Greenwich" features and are there related places, different surface forms, etc
  2. Extract list of "made in Greenwich" items
  3. Reach conclusion on terminology cross-walking between our two datasets (BM, Sci Mus) - does "Greenwich" mean anything in both or either? [Yes - we've determined that both text strings for place ultimately refer to the Borough of Greenwich but have found that this level of detail isn't enough for very meaningful browsing.]
  4. If we had more time, we'd try using http://www.w3.org/wiki/GeoOnion GeoOnion to display varying levels of precision for different records, perhaps based on how many places contained in the grid data (e.g. TQxxxx).

 

Dominic Oldman did a SPARQL query of BM CIDOC CRM data CRM:p3f .has_note made in "Greenwich"

Retrieved an item made by "Beaver" 

Object = "horse brass" which is a SKOS term x34215

Also found a UK Nat grid ref Production Place TQ4077

 

Keith then searched the OS for map resources relating to Greenwich

OS OpenData has Boundary line data available for Local Authority boundaries downloadable by Grid Square TQ

This was downloaded by email registration and would be available for plotting the data on if we had GIS available.

There is an OS API that enables you to:

 

  • Add map widgets – search box, overview map.
  • Add your own content to markers – movie links, text, images.
  • Auto marker clustering.
  • Draw routes – start/finish flags, edit routes and add circular routes.
  • Contains 1:50 000 Scale Colour Raster and OS VectorMap® District.
  • Collect the entire script for your web page.
  •  

Comments (7)

Keith May said

at 6:26 am on Dec 1, 2011

English Heritage have a map of the World Heritage Sites available via MAGIC for download
http://services.english-heritage.org.uk/NMRDataDownload/SecurePages/Download.aspx

Keith May said

at 6:27 am on Dec 1, 2011

The EH map includes the Greenwich WHS boundaries

Mia said

at 6:28 am on Dec 1, 2011

It's a bit of a tease for the ADS to link to the Ordnance Survey - I could find IDs for our location (http://data.ordnancesurvey.co.uk/doc/50kGazetteer/104227 or http://data.ordnancesurvey.co.uk/doc/50kGazetteer/104226) but didn't know how to use them to query the ADS, or even whether the ADS used them internally.

Mia said

at 6:31 am on Dec 1, 2011

Mia said

at 6:47 am on Dec 1, 2011

Mia said

at 6:55 am on Dec 1, 2011

@ostephens : @mia_out 'place' is (IMO) complicated for BM data. This is an imperfect query for items found in Greenwich bit.ly/vojTWq

Keith May said

at 3:32 pm on Dec 1, 2011

We need a better understanding of how to record and present useful metadata about Linked Data resources so users can discover what data is available for querying and re-use.
Is VoID going to be of use for this.
Do we need so protocols and agreements on what metadata is useful?

You don't have permission to comment on this page.