The document discusses the need for libraries to open up their metadata and systems to better serve their communities. It recommends that libraries (1) open their systems and data to make them more accessible, (2) actively link and share their unique data to contribute to the wider network of linked open data, and (3) experiment with new ways of sharing and discovering data through partnerships and hackathons to have broader impact.
10. “It seems to me that we’ve spent enough time on
closing down our information systems while
attempting to broaden our influence through the
production and delivery of innovative library
programs and services. When we support the
development and use of closed systems we’re
failing our communities.”
--Cecily Walker, LITA Forum 2016 Keynote
11. Open your own systems
Start to think of the library, and it’s systems as
a coherent platform, and open them
Understand the data you collect, house, and
make available, and provide clear usage
information that is both human and machine
readable
Share your data, and the information about
your data, widely
12.
13.
14. Be a network
pipe
Linked data is only going to work if
you:
◦ Actively link your own data
◦ Take ownership of the data that is
unique to you, and expose it for
consumption
◦ This means having:
◦ Namespaces
◦ Endpoints
◦ And infrastructure
19. Impact
◦ Encouraging promiscuous sharing of our data
◦ Both metadata and content
◦ Active partnering with groups like the DPLA and HathiTrust to share data broadly
◦ Seeking out interdisciplinary partners on your campuses
◦ Host hackathons, push new types of discovery
Notas del editor
Start by talking about the OSUL Discovery project and vision
Local sources:
Digital collections
Library catalog
Vended content
Website
Knowledge resources (hours, common questions, scheduling, etc)
Within the library, this seems to be accepted as an axiom. That data, in its most purest form, should be free and available to everyone. As a community, we’ve put tremendous capital behind this notion when it comes to advocating for free and open scholarly communications, for access to scholarly articles and research, to promoting open science and open data. Librarians are helping to lead discussions related to open textbooks, and open learning. We make statements that free and open data are the bedrocks of an informed and healthy democracy.
As librarians, we love open data…
Image from: https://image.slidesharecdn.com/20140722-icmeworkshop-thematerialsproject-141025221108-conversion-gate01/95/icme-workshop-jul-2014-the-materials-project-19-638.jpg?cb=1414275216
Until we don’t. We are quick to point that Libraries have silos because our vendor or our systems artificially create silos around our data. And there is some truth to that. Libraries license a lot of content. We pay people for access to information. And in those cases, those content creators or aggregators definitely create silos.
But this doesn’t let us off the hook. Libraries create silos. We love them. Early in my career – I didn’t believe this. I would look at the library, see the silos, and believe that in order for libraries to stay relevant, we needed to find ways to remove the silos from the library. But I’m not sure I believe that any more. Libraries will always have silos; it’s just the nature of being an organization that doesn’t produce all its own data. We do consume and facilitate access to a wide range of resources, from a wide range of sources. I don’t think that role will change any time soon – for either large academic libraries or for public libraries. In fact, I think that given the financial struggles that many members of our communities find themselves under, the library may be the only place where access to some silos of information will still be available.
So, this isn’t a talk about how silos are bad. I would like to see libraries find ways to make the gaps between the silos less noticeable, to create transparent bridges between the silos, but that is a very different conversation than not having silos at all.
Image from: https://methodsdigital.co.uk/wp-content/uploads/2017/01/data-silos.png
I don’t want to talk about the silos (though I will, a little later), I want to talk about us. Libraries and the communities that we build. Librarians are quick to point our the short comings of the people that we work with, the systems that we purchase and use, of the content that we license. And we should. We should be loud, and insistent and the people that we do business with demonstrate our values and allow us to fulfill our missions. But we let ourselves off the hook too quickly, and too easily.
We are very good at seeing the faults and the problems of the people around us, but as a community, as libraries, as libraries in Ohio, I think it might be time to take an honest look at ourselves and realize that as organizations, our libraries are some of the most data unfriendly organizations that we work with.
Image from: https://s-media-cache-ak0.pinimg.com/736x/17/06/73/170673457639801e22e4d13627b34b46.jpg
But wait, how can that be true? We’ve created this awesome thing called OhioLINK. We started OCLC, and then shared it with everyone else. We have companies like Zephera partnering and pushing libraries here in Ohio and beyond to “linkify” their data to make it more friendly and available to the web.
As institutions – we are doing discovery. I’d guess most people in this room have institutions that have invested in discovery projects – that use EDS, or Encore, or WorldCat Local to allow your libraries to pretend that there are no silos – that we are actually doing a good job living our values, and making data and metadata easier for our users to find and utilize.
But does any of that actually make our data, or systems, more open. Generally, I’d say no. Finding stuff or making content easier to find is definitely important, but it doesn’t make your data open, and this is a really important distinction, because humanities research and the dream of linked data in the Libraries largely depend on it.
It’s time that Libraries begin evaluating our own systems and data polities against the same criteria that we want to use when we evaluate our data partners. So what do I mean.
Well, if a researcher comes to your library and says that they would like to data mine all the full text for your institutional repository – could you give them the data? Technically, this might be difficult, but honestly, how many could say that local policy or a knee jerk “copyright” response might end the discussion. I can tell you that would likely be the case at Ohio State (I’ve asked).
Can individuals download your library catalog? Can you post your catalog online in XML for anyone to download? The answer is sure, but the reality is that most libraries are still afraid that OCLC might come after them. I can tell you that’s the case at Ohio State (again, I’ve asked).
How easy is it for libraries to utilize local control vocabularies that you may develop?
Can you tell me what your data policies are related to information that you share or data that you capture? Can I, as a user, find that information out? Can I, as a researcher, have access to the library’s anonymous search logs?
Does your library have API? Is it documented, can researchers use it?
Image from: https://fthmb.tqn.com/mlt3mdrYbvEHrsTFX_aeh6Qqj5Y=/960x0/filters:no_upscale()/about/bird-escape-97373083-resized-56a0a12b3df78cafdaa3701f.jpg
If we could all be honest, I would bet that in this room, I could find libraries that proactively shield their metadata from harvesting or reuse outside of their systems (raise my hand), libraries that attach license terms to their metadata use (raise my hand), that treat all digital content in their collections, whether in the public domain or not, as materials “owned” by the libraries (raise my hand), or purposefully shield your metadata systems (catalogs, repositories) from indexing in systems like Google (raise my hand).
I’m also sure that in this room, it would be the exception, not the rule, to find an organization that treats data the same we we’d like our content providers to make it available to us – that it’s structured, and readily available via an API. Can someone use an API to schedule rooms in your library? Can someone use an API to determine if your library is open, to ask for data in your catalog, to upload new content into a repository, to data mine your institutional repository? We regularly ask these questions and have expectations that our partners will provide many of these services, yet as organizations and communities, we routinely exclude our organizations from these requests.
And why? Because these items have value to us as organizations. They have value, because they are ours. We’ve invested our time, our treasure, or organization’s resources to create metadata and resources, and this information has value to our organization – we want credit for it…so we protect it, we hoard it.
Trust me, information doesn’t want to be free. People want other people’s information to be free…and they want their information to be valued.
Image from: https://www.njitalianheritage.org/wp-content/uploads/2015/11/pinnochio.jpg
Ok – so that is all really depressing. I’ve depressed myself a little bit; but we need to be depressed about the state of things. I actually attend a lot of meetings, and read a lot of articles patting ourselves on the back, and the reality is, if we held ourselves to the same standards that do our partners, we’d be writing a lot of articles about how much the libraries sucks.
I’m depressed because this year, I’ll be part of the group working to bring Ohio content to DPLA, and someone that is a part of this project will have to convince some libraries in this room to share their data. We will have to convince some libraries in this room that its ok to share metadata with the DPLA, and make said metadata available as public domain (no license, anyone can use it). I’m depressed because we will hear, maybe from some libraries in this room, that this isn’t possible, or things can’t be shared for many of the same reasons that many of our content vendors give to us, and no one will see the irony.
I’m depressed because my largest communities, OhioLINK and OCLC make it almost impossible, as a researcher, to use my own organization’s data – let alone the rich community data that I’d love to get my hands on. I’m depressed because I look at the work that is involved if libraries want to do linked data, and realize that this work can’t happen until libraries understand that linked data doesn’t mean just putting links in your metadata – it means being a data source for other people. And if making our data available makes us uncomfortable now – the linked data world will isn’t going to feel good at all.
Image from: http://static.boredpanda.com/blog/wp-content/uploads/2015/06/11192909_669167436562807_839335436_n.jpg
Let’s start here. Libraries are at the for front in open education and open learning. We are developing new services, and finding ways to build new partnerships. We are doing our best work for the users that we can see, for the needs that we can identify and answer right now.
It’s the users that we don’t see, and the opportunities lost where we are failing right now. I’m asked all the time, how can libraries make themselves ready for linked data. My answer…open your data and think of the Library as a platform, and your data as a its service.
Quote taken from Code4Lib 2017 Keynote, Christine Harlow
Never in my life has memos and note taking been such a hot topic in the news. But as metadata folks, as the one’s tasked with building open systems, we need to start documenting our decisions, and making those decisions open to the public.
Open data and open systems are not about public domain, or open source licenses (though there is some of that), but its about transparency, honestly, and trust.
For all of it’s flaws – WikiData does a good being transparent. From this page, you can see not only how to get the data, but information about various interfaces, and best practices.
It’s time to get off the side-lines and start experimenting with our data.
Additionally, we need to open up our data for experimentation.