Posting to NGC4LIB
On 10/10/2013 5:15 PM, Karen Coyle wrote:
Jim, before you throw out the baby with the bathwater, consider that the linked data universe is a two-way street, and some folks are starting to take advantage of that. Try some searches in the linked catalog at FAO:
http://aims.fao.org/openagriswhere data is pulled into the display that provides supporting information for the actual article or report being displayed, based on data in that article.
And try a search at:
https://apps.facebook.com/worldcat/and when you get your results, click on “related people and topics”. That shows you what links you get using the linked data that is in WorldCat today, and that’s without identifiers for places or other types of data that are available for linking.
I don’t know what your vision of linked data is, but it can provide simple services like book covers and author bios, or it can be big and complex. In some instances it may just be a standardized replacement for some of the APIs people use today. In others, like the use that is made of it by the BBC web site, it really becomes a way to generate whole pages of information around a topic. Don’t think of it as a library standard — it’s a data standard, and it’s being heavily used as such. It doesn’t have to dominate the library view of its data, but we would be foolish not to make use of it.
I have said repeatedly that libraries should go for linked data. The major problem is the idea that going into linked data will solve any of the real challenges facing libraries. While linked data is a promising technology, it is still very young and history is littered with the remains of long-forgotten “promising technologies”. Libraries, of anybody out there, should understand that very well since so often in the past many libraries have bet on technologies that seemed promising but ended up in the junkyard. Plus, there are lots of promising technologies on the web today in addition to linked data and new ones are popping up all the time.
Yet, linked data is indeed promising. So, I have said that libraries should enter the linked data universe, but do it in the least disruptive and cheapest way. Doing so does not demand new cataloging rules, RDA, FRBR-type formats and so on. FAO doesn’t have any of that and they are being held up as a model. So, if that’s true, then do what FAO did and enter into it in the least disruptive and cheapest ways possible. You do not do this by changing rules and cataloging systems. You do it by exporting a certain type of format and making that format available. There are many ways of doing that. Luckily, you don’t even have to create full-scale RDF today, and there are far easier methods.
But this is not the way the library world is going. They have convinced themselves that RDA is necessary because the public wants it. (Really??!! Who exactly?) That FRBR-style records are needed to enter linked data. (Really??!! What about everybody who is in there already without it?) Going the library route has been expensive already, and will promise to be much more costly in the future. Not exactly a recipe for success today.
But nevertheless, go into it with a minimum of fuss and bother. That’s fine. I have no doubt that once we are in the linked data universe and have gained some experience, seeing how positively (or negatively) the public and webmasters react, then we will have some real information to base future decisions on. I am sure we will discover a lot that we do not want to know, but I am just as sure that we will find we need to change some cataloging rules and formats and all kinds of other things.
Spending fabulous amounts of money, while waiting even longer before moving into the linked data cloud, all in the faith that it promises some kind of salvation, is the equivalent of betting the baby’s shoes on a dice roll. You may get lucky and roll a 7, or you may be looking at “snake eyes”.