The morning was MCd by Costis Dallas and there was a welcome and introduction from the chair of the Europeana Foundation before Jill Cousins (Europeana Foundation) provided an overview of Europeana. I'm sure the figures will be available online, but in summary, they've made good progress in getting from a prototype in 2008 to an operational service in 2010. [Though I have written down that they had 1 million visits in 2010, which is a lot less than a lot of the national museums in the UK though obviously they've had longer to establish a brand and a large percentage of their stats are probably in the 'visit us' areas rather than collections areas.]
Europeana is a super-aggregator, but doesn't show the role of the national or thematic aggregators or portals as providers/collections of content. They're looking to get away from a one-way model to the point where they can get data back out into different places (via APIs etc). They want to move away from being a single destination site to putting information where the user is, to continue their work on advocacy, open source code etc.
Jill discussed various trends, including the idea of an increased understanding that access to culture is the foundation for a creative economy. She mentioned a Kenneth Gilbraith [?] quote on spending more on culture in recession as that's where creative solutions come from [does anyone know the reference?]. Also, in a time of Increasing nationationalism, Europeana provided an example to combat it with example of trans-Euro cooperation and culture. Finally, customer needs are changing as visitors move from passive recipients to active participants in online culture.
Europeana [or the talk?] will follow four paths - aggregration, distribution, facilitation, engagement.
- Aggregation - build the trusted source for European digital cultural material. Source curated content, linked data, data enrichment, multilinguality, persistent identifiers. 13 million objects but 18-20thC dominance; only 2% of material is audio-visual [?]. Looking towards publishing metadata as linked open data, to make Europeana and cultural heritage work on the web, e.g. of tagging content with controlled vocabularies - Vikings as tagged by Irish and Norwegian people - from 'pillagers' to 'loving fathers'. They can map between these vocabularies with linked data.
- Distribution - make the material available to the user wherever they are, whenever they want it. Portals, APIs, widgets, partnerships, getting information into existing school systems.
- Facilitate innovation in cultural heritage. Knowledge sharing (linked data), IPR business models, policy - advocacy and public domain, data provider agreements. If you write code based on their open sourced applications, they'd love you to commit any code back into Europeana. Also, look at Europeana labs.
- Engagement - create dialogue and participation. [These slides went quickly, I couldn't keep up]. Examples of the Great War Archive into Europe [?]. Showing the European connection - Art Nouveau works across Europe.
The next talk was Liam Wyatt on 'Peace love and metadata', based in part on his experience at the British Museum, where he volunteered for a month to coordinate the relationship between Wikipedia as representative of the open web [might have mistyped that, it seems quite a mantle to claim] and the BM as representatiave of [missed it]. The goal was to build a proactive relationship of mutual benefit without requiring change in policies or practices of either. [A nice bit of realism because IMO both sides of the museum/Wikipedia relationship are resistant to change and attached firmly to parts of their current models that are in conflict with the other conglomeration.]
The project resulted in 100 new Wikipedia articles, mostly based on the BM/BBC A History of the World in 100 Objects project (AHOW). [Would love to know how many articles were improved as a result too]. They also ran a 'backstage pass' day where Wikipedians come on site, meet with curators, backstage tour, then they sit down and create/update entries. There were also one-on-one collaborators - hooking up Wikipedians and curators/museums with e.g. photos of objects requested.
It's all about improving content, focussing on personal relationshiips, leveraging the communities; it didn't focus on residents (his own work), none of them are content donation projects, every institution has different needs but can do some version of this.
[I'm curious about why it's about bringing Wikipedians into museums and not turning museum people into Wikipedians but I guess that's a whole different project and may be result from the personal relationships anyway.]
Unknown risks are accounted for and overestimated. Unknown rewards are not accounted for and underestimated. [Quoted for truth, and I think this struck a chord with the audience.]
Reasons he's heard for restricting digital access... Most common 'preserving the integrity of the collection' but sounds like need to approve content so can approve of usages. As a result he's seen convoluted copyright claims - it's easy tool to use to retain control.
Derivative works. Commercial use. Different types of free - freedom to use, freedom to study and apply knowledge gained; freedom to make and redistribute copies; [something else].
There are only three applicable licences for Wikipedia. Wikipedia is a non-commercial organisation, but don't accept any non-commercially licenced content as 'it would restrict the freedom of people downstream to re-use the content in innovative ways'. [but this rules out much museum content, whether rightly or not, and with varying sources from legal requirements to preference. Licence wars (see the open source movement) are boring, but the public would have access to more museum content on Wikipedia if that restriction was negotiable. Whether that would outweight the possible 'downstream' benefit is an interesting question.]
Liam asked the audience, do you have a volunteer project in your institution? do you have an e-volunteer program? Well, you do already, you just don't know it. It's a matter of whether you want to engage with them back. You don't have to, and it might be messy.
Wikipedia is not a social network. It is a social construction - it requires a community to exist but socialising is not the goal. Wikipedia is not user generated content. Wikipedia is community curated works. Curated, not only generated. Things can be edited or deleted as well as added [which is always a difficulty for museums thinking about relying on Wikipedia content in the long term, especially as the 'significance' of various objects can be a contested issue.]
Happy datasets are all alike; every unhappy dataset is unhappy in its own way. A good test of data is that it works well with others - technically or legally.
According to Liam, Europeana is the 21st century of the gallery painting - it's a thumbnail gallery but it could be so much more if the content was technically and legally able to be re-used, integrated.
Data already has enough restrictions already e.g. copyright, donor restrictions. but if it comes without restrictions, its a shame to add them. 'Leave the gate as you found it'.
'We're doing the same thing for the same reason for the same people in the same medium, let's do it together.'
The next sessions were 'tasters' of the three thematic tracks of the second part of the day - linked data, user-generated content, and risks and rewards. This was a great idea because I felt like I wasn't totally missing out on the other sessions.
Ian Davis from Talis talked about 'linked open culture' as a preview of the linked data track. How to take practices learned from linked data and apply them to open culture sector. We're always looking for ways to exchange info, communicate more effecively. We're no longer limited by the physicality of information. 'The semantic web fundamentally changes how information, machines and people are connected together'. The semantic web and its powerful network effects are enabling a radical transformation away from islands of data. One question is, does preservation require protection, isolation, or to copy it as widely as possible?
Conjecture 1 - data outlasts code. MARC stays forever, code changes. This implies that open data is more important than open source.
Conjecture 2 - structured data is more valuable than unstructured. Therefore we should seek to structure our data well.
Conjecture 3 - most of the value in our data will be unexpected and unintended. Therefore we should engineer for serendipity.
'Provide and enable' - UK National Archives phrase. Provide things you're good at - use unique expertise and knowledge [missed bits]... enable as many people as possible to use it - licence data for re-use, give important things identifiers, link widely.
'The goal is not to build a web of data. The goal is to enrich lives through access to information.'
Kevin Sumption from the National Maritime Museum presented on the user-generated content track. A look ahead - the cultural sector and new models... User-generated content (UGC) is a broad description for content created by end users rather than traditional publishers. Museums have been active in photo-sharing, social tagging, wikipedia editing.
Crowdsourcing e.g. - reCAPTCHA [digitising books, one registration form at a time]. His team was inspired by the approach, created a project called 'Old Weather' - people review logs of WWI British ships to transcribe the content, especially meterological data. This fills in a gap in the meterological dataset for 1914 - 1918, allows weather in the period to be modelled, contributes to understanding of global weather patterns.
Also working with Oxford Uni, Rutherford Institute, Zooniverse - solar stormwatch - solar weather forecast. The museum is working with research institutions to provide data to solve real-world problems. [Museums can bring audiences to these projects, re-ignite interest in science, you can sit at home or on the train and make real contributions to on-going research - how cool is that?]
Community collecting. e.g. mass observation project 1937 - relaunched now and you can train to become an observer. You get a brief e.g. families on holidays.
BBC WW2 People's War - archive of WWII memories. [check it out]
RunCoCO - tools for people to set up community-lead, generated projects.
Community-lead research - a bit more contentious - e.g. Guardian and MPs expenses. Putting data in hands of public, trusting them to generate content. [Though if you're just getting people to help filter up interesting content for review by trusted sources, it's not that risky].
The final thematic track preview was by Charles Oppenheim from Loughborough University, on the risks and rewards of placing metadata and content on the web. Legal context - authorisation of copyright holder is required for [various acts including putting it on the web] unless... it's out of copyright, have explicit permission from rights holder (not implied licence just cos it's online), permission has been granted under licensing scheme, work has been created by a member of staff or under contract with IP assigned.
Issues with cultural objects - media rich content - multiple layers of rights, multiple rights holders, multiple permissions often required. Who owns what rights? Different media industries have different traditions about giving permission. Orphan works.
Possible non-legal ramifiations of IPR infringements - loss of trust with rights holders/creators; loss of trust with public; damage to reputation/bad press; breach of contract (funding bodies or licensors); additional fees/costs; takedown of content or entire service.
Help is at hand - Strategic Content Alliance toolkit [online].
Copyright less to do with law than with risk management - assess risks and work out how will minimise them.
Risks beyond IPR - defamation; liability for provision of inaccurate information; illegal materials e.g. pornography, pro-terrorism, violent materials, racist materials, Holocaust denial; data protection/privacy breaches; accidental disclosure of confidential information.
High risk - anything you make money from; copying anything that is in copyright and is commercially availabe.
Low risk - orphan works of low commercial value - letters, diaries, amateur photographs, films, recordings known by less known people.
Zero risk stuff.
Risks on the other side of the coin [aka excuses for not putting stuff up]