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

  • Browse and search Google Drive and Gmail attachments (plus Dropbox and Slack files) with a unified tool for working with your cloud files. Try Dokkio (from the makers of PBworks) for free. Now available on the web, Mac, Windows, and as a Chrome extension!


Linking Museums II: write-ups

Page history last edited by Mia 10 years, 10 months ago

Notes from the Linking Museums II meetup on September 27, 2010, London.


Initial (vague) notes by Mia but please feel free to edit and add your notes, or comment on the page if that's easier for you.

Who was there?

James (Kew), @jamesinealing

Jonty @Jonty

Paul Miller @PaulMiller

Greg Hadfield @greghadfield

Richard Boulton @rboulton

Joe Padfield (National Gallery) @JoePadfield

Jo Pugh (Nat. Archives) @mentionthewar

Jim O'Donnell @pekingspring

Mia (Science Museum), @mia_out


What was discussed?

  • @leifuss couldn't make it but had been in contact earlier to suggest Antiquist could offer a prize to the first cultural heritage organisation to offer actual linked, open data (as his initial survey findings were that no cultural heritage organisation was there yet).  What do you think?
  • Paul mentioned a relevant JISC call, out in a few weeks
  • We did a round of introductions (partly because I am selfish and wanted a chance to eat my dinner). Quite a few people from cultural heritage organisations were there because they are interested in publishing their data through APIs or as structured data and weren't sure where to start.  And of course some developers were there because they want access to our data, hoorah!
  • People at the start of the process of publishing data mentioned some of the issues they were encountering, including data sharing concerns from staff - attribution.  Also, fear of/perceived penalties for being wrong (well, more precisely - being wrong in public).
  • First question from developers when someone mentioned releasing data - what licence would it be released under?
  • Joe has been charging ahead and is close to having entire collection online and linked to dbpedia.  Web with SPARQL endpoints.  Queries dbpedia to see if artists are in it.  Worked on date range data for paintings/artists. CIDOC-CRM, has-possible, has-probable (fuzziness). Dealing with 'artist active' dates vs 'artist lived' dates - sometimes only know when they were active, not when they were born or died.  [Apologies for the mangled notes, please feel free to correct them]
  • Mia is looking to put subject authorities (people, places, events, topics) online so would love to work with people on testing the usability of the data for developers (alongside the user experience for the web visitor).
  • Comment: GLAM data needs pictures to be compelling.
  • Comment: 'a hack day with dull data is no fun for anyone'.
  • 'labs' areas are really useful for working on projects in public for feedback and input and seems to be a way for people to understand 'beta', pilot-ish nature of the work and manage expectations around that
  • Joe: D2RQ - can use to map collections database to live SPARQL endpoint or produce a dump file. [Cool!]


Where should this stuff live?

Is the wiki the best place for people to share stuff, think aloud, get feedback and work out what 'best practice' looks like?  What do you think?

Related: what's the best way to push notifications at people - twitter, mailing list?  What do you think?


In more detail - where should these discussions and beta work live?  I know the wiki could do with a tidy-up generally, but we could also move to another platform if that'd be more usable.  Joe has stuff on another, password-protected wiki about work he's done that he could put somewhere public, I'd love to share bits as I'm working on them and there are already some pages like that on the wiki.  Ideally whatever platform we use should be discoverable by search engines, and content should include attribution. 


Concrete issues for discussion include: dealing with dates, dealing with subject authorities (as discussed); a 'how to' for content providers (look also at what spacetimecamp and other hack days, open/linked data projects have produced, also Good APIs project, info from Europeana, etc?); a list of resources people have found useful; lists of questions to ask when assessing formats, standards; how to choose a licence that matches your existing terms and test it against proposed usage, etc...


Provocative questions

Some really useful questions were asked or emerged as issues over the evening - I'd love to know what you think.


  • James: can you provide tools to help non-tech people use your data? Further, on discussion, what would an interface that helped people help themselves be like?
  • Mia: as developers, what do you want from a subject authority? [a museum-y term, I know, but it's people, organisations, places, events, etc. e.g. biographies of scientists, events or places relevant to our collections]
  • Is there any value in a survey of developers that would help museums understand the type and level of demand for machine-readable data?  Ask developers their top wish request from GLAM (galleries, museums, libraries and archives).  It could help prove demand (or not) and help museums understand what users need and better direct efforts. It could be as simple as: 1. what do you want to do with the data? 2. How do you want to access it? 3. what rights do you need? (variation: do you want to use it commercially?)
  • Mia: geek accessibility - is the mention of RDF/SPARQL to a mid-level or tinkerer developer what stairs are to Daleks?
  • [can't remember who]: what if the British Museum was a start-up?
  • What does a good request for data look like? AKA, moving beyond FOI to sustainable (for the org) reusable (for the dev) provision of data. 


Other stuff

Other events - http://historyhackday.pbworks.com

Next meetup

Possibly at spacetimecamp meetup on November 5th, and then maybe one later in November? Jonty to organise (yay!).

Comments (5)

Owen Stephens said

at 10:30 am on Sep 30, 2010

Two comments stand out to me:

Comment: GLAM data needs pictures to be compelling.
Comment: 'a hack day with dull data is no fun for anyone'

I think this is drastically underestimating the potential interest in data. To be honest it is about having ideas about how we manipulate, enhance and present data - and this doesn't necessarily mean having pictures or shiny things. Bibliographic metadata isn't terribly visual and for most intents and purposes I think we could say it was 'dull' in an everyday sense - and yet each Mashed Library (http://www.mashedlibrary.com/) event produces interesting data hacks and people have fun. In the recent 'Hack Warwickshire' comptetition based around data released by Warwickshire County Council, one of the (2) winning entries was based around an RSS feed of 'new books in the library' - and done by a non-librarian.

Let's not underestimate the interest there is in our data :)

Richard Boulton said

at 2:55 pm on Sep 30, 2010

I think I may be the source for "GLAM data needs pictures to be compelling", though I meant to be a bit more nuanced than that!

Pictures aren't needed, but they do help to make a lot of potential uses of the data far more interesting to users.

Even for somewhat abstract visualisations it's really good to be able to drill down through the stats to get details of individual objects, people or whatever, and pictures help to make those details much more tangible and understandable. For example, imagine a project to explore the development of scientific techniques by linking scientists to the experimental apparatus they used. A network or timeline view of that data could be very interesting by itself, but would be far more compelling if it were possible to display pictures of the apparatus (and the scientists)!

I certainly wouldn't want to stop anyone releasing data just because they didn't have pictures to go along with it. My concern is more that, since there's often licensing problems with providing pictures, there may be a temptation to release without pictures which would make that data far more useful and interesting. I'd like to encourage those releasing data that developers will value pictures released along with data highly, and we'll appreciate the effort taken to release them.

Mia said

at 9:00 pm on Oct 4, 2010

That'd be my quick and dirty summaries at fault, sorry!

The 'needs pictures to be compelling' comment wasn't mine, but I can see that object records have more potential to be interesting when they have images.

Phill Purdy said

at 11:49 am on Oct 7, 2010


I’d like to extend an invitation to the group to the first ever Culture Grid Hack Day and offer it as an opportunity for a get-together in Newcastle: http://www.culturegrid.org.uk/use/hack-day/

Best, Phill

Mia said

at 1:22 pm on Oct 8, 2010

Thanks Phill - unfortunately I won't be able to make it, but it's prompted me to add a page for events to this wiki: http://museum-api.pbworks.com/w/page/Events

'Tickets' for the spacetimecamp meetup on Nov 5 are now available... http://spacetimecamp.eventbrite.com/

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