Posting to Bibframe
On 01/02/2017 22:25, Jeff Edmunds wrote (concerning the definition of “failure”):
Widely adopted = adopted by a majority of libraries (of which, in the US, according to ALA, there are approximately 119,000, the vast majority of which use MARC now)
Just as important as a definition of failure is a definition of success. Success here seems to be equated with getting a majority of libraries to use Bibframe. If we are considering Bibframe as a path to some future linked data universe, then this is certainly different from the original dream of linked data. Simply put, the original dream of linked data is to put your data on the web in a coherent way for others, i.e. those who do not know or understand your data, to use it for their purposes. The idea of linked data is not that you can do something new with your own data–after all, you already understand how your own data is structured and you have complete control over it to do whatever you want. It’s for others outside of your own community to use your data as they want.
Tim Berners-Lee has given lots of examples, e.g. a quote: “A classic story, the first one which lots of people picked up, was when in March — on March 10th in fact, soon after TED — Paul Clarke, in the U.K. government, blogged, “Oh, I’ve just got some raw data. Here it is, it’s about bicycle accidents.” Two days it took the Times Online to make a map, a mashable map — we call these things mash-ups — a mashed-up user interface that allows you to go in there and have a look and find out whether your bicycle route to work was affected.” (From: http://www.ted.com/talks/tim_berners_lee_the_year_open_data_went_worldwide. If you haven’t already seen this talk, I suggest it)
If Bibframe is supposed to be aimed only toward libraries, then that means it is aimed at members of the same community who already understand how the data is structured. With such a group, very similar results could be achieved with OAI-PMH, APIs, or even crosswalks, depending on what someone wants to do. That would be much simpler, cheaper, and faster. Still, that means building new services and of course, that requires lots of time, is expensive and involves risk. We should never forget that the garbage dump of the web is littered with services that people have rejected. (Remember MySpace? I found out it still exists!) There is absolutely no guarantee that if libraries make new services, that anyone will like them one bit more than our current catalogs.
If the purpose of Bibframe is to “open the silos” and put our metadata where the users are, that would mean (to me at least) that libraries want their data included in the search results of Google, Bing, Yandex, perhaps even Baidu, and other search engines that are used by the overwhelming majority of the public. In that sense, the risk mentioned above about creating new services is avoided. But to get your data into the search engines you must use schema.org. Not Bibframe. Enough said. If you want your data on Facebook, it’s got to be Open Graph.
Of course, even when you get your data in the popular search engines your problems are only beginning: it’s true that your linked data and your structures tend to disintegrate, but most important: since there will be no links to your data, it will remain trapped at the very bottom of the search engine results, much like sludge at the bottom of an oil tank. Raising those results has proven to be incredibly difficult and expensive, and even then it doesn’t always work. It is possible that in the future, those search engines may change their policies and allow Bibframe, but the chances for that seem to be very, very low.
If we expect webmasters to use our library data to create new tools such as those described by TBL, there is no guarantee that anyone will want Bibframe data, especially if it is very complicated for them to do so. Webmasters would prefer to take their bibliographic data from easier sources, e.g. Amazon or Google, as they already do and where they don’t have to pay for it (open). Promoting our bibliographic information as being “superior” would be largely meaningless to most of them, I suppose, especially if our data is not open (if they would have to pay for it).
Yet, I believe the public (and by extension, webmasters) would absolutely love some of the data of libraries, primarily circulation data, so that people could discover, e.g. what are the most popular books checked out by Ivy-League undergraduates, or undergraduates in London. They would love to know what is checked out by graduate students in business schools, or by senior faculty in political science, and so on. I would like that myself! Instructors designing course syllabi would love to know what are the assigned readings for similar courses taught elsewhere (i.e. what is currently on Reserve). But letting out circulation information has serious ethical consequences for librarians, and most of it is not part of Bibframe anyway.
In this scenario, which I consider describes the real world, I don’t know how to measure success and failure. The Bibframe initiative should be looking forward toward how to deal with such highly obvious issues. It is not a matter of simply believing in: “Build it and they will come!” That is a sure-fire recipe for disaster.
I think it would be safe to assume that if nothing real is produced for the public in the next 5-10 years, it will be seen as a failure? Of course, the public would still have to be convinced that what we make is wonderful and administrators would need to be convinced it has an adequate ROI …