Over the course of the two-day NISO BIBFRAME Roadmap Meeting last week, I kept circling back to thinking about our goals. There was a discussion subgroup that focused on Goals and Strategies, and they raised some very important issues, but I didn’t come away feeling that we had a strategy for addressing them. There were a lot more questions identified than solutions, and the goals remained nebulous and abstract.
I’d like to delve a little deeper into some of the goals that were identified at the meeting, and to pull apart what I think are two different focal areas. There are the goals behind creating a new data model, and what we are trying to achieve with it, and then there are the goals of implementation of a new data model, and what it will take to move beyond ideas and into practice. I think these might be difficult to pick apart, but I want to try to keep them separate in my mind, and try not to conflate them.
There were several high level goals identified at the meeting and I’d like to talk about them here, and begin to brainstorm some more actionable goals and concrete strategies around them. These goals are very closely intertwined, and they are really about what a new data model would be able to achieve. There were many more goals identified, but these are the five I found to be most compelling (and focused on the model itself, not an implementation plan).
1) To improve interoperability. As I wrote in my last post on MARC, one of MARC’s major flaws is the lack of interoperability outside of the library community. Even within library systems it can be challenging to bring information about our different resources together, hence, the frequent lamentations about how silo-ed our metadata is. One crucial, big-level goal for a new data model is that it needs to be able to be used in a variety of different systems, using modern data and web technologies.
2) To enable use by a broader community. This is closely tied to the first goal. Right now, our rich bibliographic data and our valuable holdings information lives in a series of walled gardens. It’s hard for non-library developers to access and use this metadata systematically. In moving towards a new framework, we should be aware of external communities that have an interest in bibliographic data and library holdings data, communities we don’t even know about yet. We should ensure that methods for accessing library data are consistent and transparent.
3) To advocate for and endorse open metadata policies. In order for our metadata to be used by other communities, and shareable within our own, we need to have clear policies of openness. Right now, there is uncertainty in the library community about who owns what data and what we have a right to do with it. But openness will be crucial for the first two goals to be meaningfully achieved, and it is a goal in its own right.
4) To enhance collaborative and shared infrastructures. Right now, libraries have a wonderful infrastructure for sharing data among ourselves, but it could be even better. A new model should improve our systems for cataloging, resource sharing, and discovery. It should improve our ability to share metadata with each other. It should not erase or dismantle the networks that we’ve spent so much time creating, but should take advantage of and improve them.
5) To ensure better visibility of library data for end users on the web. This is, to me, the one goal to rule them all. I believe all the other goals are in service to this: Making sure that our users can find and use our resources. And that means not just on the web as we know it today, but on the web that will exist tomorrow. Any new data model has to make it easier for users to discover what we have, and to understand how they can use and access it. And it needs to happen outside of our library systems.
With these high-level goals in mind, what strategies might we adopt to push them forward? Looking back through my notes from the meeting, and from the notes that Nettie Lagace and the other meeting participants put together, I pulled out some concrete activities that would advance these goals, and help us answer some of the questions that were raised.
First, we need to assess the weaknesses of our current systems for shared cataloging, resource sharing, resource discovery, and user access. What things are consistently difficult for our staff? For catalogers? For users? What would improve our workflows? What would improve our systems?
Second, we need to create prototypes for how catalogers and metadata librarians would produce linked data from scratch. What would a new cataloging tool look like? Related to this, we need better prototypes for the effective migration of existing data. We also need to take a hard look at what elements in our existing metadata need to be maintained to ensure continued functionality for librarians and library users.
We should also assess current policies governing our metadata use, and advocate with our partners for open metadata policies, and for open access to metadata. What policies affect our local metadata? How do policies of different systems and vendors effect each other? Are we clear on what we can and can’t share, and if we can’t share, how can we begin to make changes to ensure that we can in the future?
These are still not specific, measurable, or time-bound goals (although I do think they are relevant and attainable). But they are somewhere more concrete where we can start. If we’re going to do the hard work of re-thinking and re-developing our existing metadata infrastructure, we should make sure the answers to some of these questions have been addressed. Right now, I get the sense that some people feel the answers are clear, but not everyone agrees.
I would love to hear your thoughts about some of these goals and strategies. Are there things you already know about some of the questions I’ve raised? Do you have some ideas about how to make these ideas into better, more concrete actions? Do you think I’ve missed (or misunderstood) a key goal for a new data model? Let me know!