These are my running notes from day 3 of the Museums and the Web conference – as the perfect is the enemy of the good I'm getting these up 'as is'. I did a demo [abstract] in the morning but haven't written up my notes yet – shame on me!
The session 'Building and using online collections' included three papers, I've got notes from all three but my laptop battery died halfway through the session so only some of them are already typed – I'll update this entry when I can sneak some time.
Paul Rowe presented on NZMuseums: Showcasing the collections of all New Zealand museums (the linked abstract includes the full paper and slides).
4 million NZers, 400 museums. NZMuseums website – focal point for all NZ museums. NSTP administers the site, Vernon Systems is solution provider.
Each museum has a profile page including highlights of their collections. Web-based collection management system.
What needs to be in place for small museums to contribute? How can a portal be built with limited resources? What features of the website would encourage re-use of the data?
Some museums had good web presences, but what about the small museums? Facing same issues that small or local govt museums in the UK face.
Museums are treasures of the country, they show who we are. Website needs to reflect that.
Focus groups – volunteers are important – keep it simple; keep costs low; some places had limited internet connectivity; reservations about content being on the internet were common.
Promoting involvement to the sector – used existing national monthly newsletters to advertise workshops and content deadlines. Minimum of 20 items for placement on site to avoid 'box ticking' [some real commitment required]. Used online forum for FAQs.
Lack of skills – NSTP were trained so could then train staff and volunteers in museums. Digitising, photography for the web.
Had to explain benefits to small museums. It gave them an easy start to getting an online presence.
They overcame resistance by allowing watermarking and clear copyright statements; they showed existing museums sites that allowed tagging; promoted that would help them reach a diverse dispersed audience.
First tag on site – 'shiny nose'. First comment was someone admitting they'd touched the nose on a bronze sculpture.
Could also import Excel spreadsheets as content management system didn't exist at early stage of project. Also provided a workaround for people with lack of internet – the spreadsheet could be posted on CD.
API provides glue to connect eHive (Collections Management System) and NZMuseums site together.
Tips for success
Use OS software where possible; use existing online forums and communication networks to save answering questions over again.
90% of these collection items not previously available on the internet. 99% of collection items have images.
[Kiwis are heroes! Everyone was incredibly modest about their achievements, but I think they're amazing.]
Next was Eero Hyvönen on CultureSampo – Finnish Culture on the Semantic Web 2.0: Thematic Perspectives for the End-user (the linked abstract includes the full paper and slides).
Helsinki semantic web thingies
Part of national ontology project, Finland
Vision – international semantic web of cultural heritage. Marriage between semweb and web 2.0
Challenges – content heterogeneity, complexity
Other challenge relates to the way cultural content is produced – Freebase, Wikipedia, open street maps, etc,
Semweb for data integration; web.2 0 approach for content production
Automatically enriched by each piece of knowledge.
In Finnish the sampo is a magic drum that makes everything possible.
Portal intended for human users and machines. Trying to establish a national way of producing content so can be published automatically.
Infrastructure – 37,000 class concepts in ontology. MAO, TAO – museum ontologies, collaboratively built ontologies, then mapped to national system. End user sees one unified ontology. [A little pause while I pick my jaw up from the ground.] 66 vocabularies, taxonomies and ontologies available online as services, can be used as AJAX widgets. Some vocabularies are proprietary so can't be published online in the service.
28 content providers, 22 libraries and museums and some international associates like Getty places, Wikipedia.
16 different metadata schemas. [Including some for poetry!]
134,000 cultural collection items (artefacts, books, videos, etc)
285,000 other resources (places, people etc)
Annotation channel for content items – web 2.0 type interface.
Semantic web 2.0 portal
Portal users – for humans, Google-like but semantic search. Nine perspectives into cultural heritage. Three languages. Recently view items, recently commented items.
[Sadly my laptop died here and the rest of my notes are handwritten. You can probably get the gist from the published paper and the slide, but the coolness of their project was summed up by this tweet: Musebrarian: What can you do with a semantic knowledgebase? Search for "beard fashion in Finland" across time and place. #mw2009
It might not sound like much, but the breadth of content, and the number of interfaces onto it was awe-inspiring.]
Sadly my notes from Brian Dawson's paper, Collection effects: examining the actual use of on-line archival images are also still on notepaper. The paper was a really useful examination of analytical approaches to understanding the motivations of people using cultural heritage collections.