On 22/02/2012 21:43, Kevin M Randall wrote:
I recently came back from an excellent NISO/DCMI webinar presentation by Karen Coyle on linked data called "Taking Library Data From Here to There" (which I highly recommend). Karen used William Shakespeare's "As You Like It" as an example in one part, and that got me thinking about a way to use the same title in a response to the above paragraph.
Freedom in using the FRBR WEMI entities in OPAC design is virtually limitless. *One* possible way that I could imagine off the top of my head is, in response to a search for William Shakespeare's "As You Like It", the user gets something that indicates that work. Maybe it's a picture of the cover of a printed edition of the play, maybe it's a bunch of words--just something to get across to the user the concept of William Shakespeare's "As You Like It". Then, accompanying that, there might be buttons or links that lead to various things:
Those are resources that are typical of what we currently describe in our metadata. But with linked data that extend outside the "library database" most of our catalogs are currently limited to, we might be able to add buttons or links leading to:
- printed editions of the play
- translations of the play
- film and TV adaptations of the play
- resources that critique, analyze, or are otherwise about the play
- other resources that relate in some way to William Shakespeare or "As You Like It"
- information on theatrical productions of the play, including reviews, performance schedules, ticket sales, etc.
- IMDB articles about films of the play
- YouTube videos of the play
- discussion forums about the play or William Shakespeare
All this is fine, and they suggest a few of the possibilities of entering the linked data world but will still demand a *ton* of work to generate correct links and/or correct queries. Librarians certainly cannot build something like this on their own but will have to have a massive amount of cooperation from other non-library agencies. This will be a new experience where catalogers will discover they are not in control at all. Still, it is absolutely necessary if we are to enter that world.
I have nothing against any of this. At the same time, I have no doubt that other agencies have not and will not adopt our FRBR structures to enter the linked data world. So, if the ultimate goal is for us to enter the linked data world, why do we have to adopt the RDA/FRBR record structure first? Why not do just do it now? "All"(!) we have to do is open our data, provide it in a format that others can use, and link what can be linked. Easier said than done, but certainly far easier than redoing our cataloging rules and waiting x number of years until FRBR is completed. One place I worked at and helped create the format has gotten into linked data using an XML application profile. Microdata can apparently do it too and it is a lot easier. http://schema.org/docs/gs.html Anyway, this is one reason why I have said that the absolute, #1 first step is to get rid of our ISO2709 format for sharing records, or at the very, very least, abandon the requirement of "roundtripability". It will have to be done eventually, and nobody can do anything with our data until that is accomplished.
Perhaps some methods are theoretically better than others in the linked data world, but we should accept that nothing is perfect. The main thing is to enter it first and then work out the problems. FRBR and its entities are not necessary to enter the linked data world. What is necessary is to provide a format (probably MODS would be good enough), put in some links where they exist, e.g. into VIAF and id.loc.gov (better than nothing) and most important: open our data.
But as I mentioned before: easier said than done.