On 23/02/2015 20.58, Gene Fieg wrote:
> While doing some updating of authorized personal names here at home, I have > noticed that some authority records are coded as rda, but very few, > sometimes none, of the fields are used even when the data is in record. > Such data as the birth and death dates, affiliation, gender, etc. >
> If those fields are not included but the info is readily available, then > all the talk about linked data is empty. Just empty. But if we are > serious about it, let’s get on the bus and do the work when it comes of > “rda’ing” authority records.
It is important not to confuse RDA with linked data. They are separate and do not need one another. There are many sites that are fully compatible with linked data right now that have nothing to do with RDA–and never will. RDA can also exist without linked data, quite easily.
The idea of linked data is *not* to copy all information into a single “mother of all databases,” then put it on the web in some linked-data-friendly format such as BIBFRAME, and then expect everyone to use what you have made. Rather, the idea of linked data is to put the information you have on the web in specific ways so that others can leverage the information you have, taking what they want and ignoring other parts, and in this way make it easier for everybody. In return, you can then use information on other sites for your own purposes. As I have pointed out earlier: the purpose of linked data is *not* to make your data obsolete! It is to enhance everyone’s data.
Sites that utilize linked data can make use of information in Wikipedia or Google Maps and do not have to copy anything from Wikipedia or Google Maps. Linked data sites just add the link(s) to the relevant information they want, and then the machines work their magic. Therefore, in a correctly built system, if information is in Wikipedia/dbpedia/Wikidata conglomeration, it is available to you just as easily as if it was in your own database. This saves everyone from copying everything over and over and over …, which has always been a huge waste of resources and effort. (Or at least it has been portrayed that way)
As an example, I just found this site: http://www.pw.org/literary_magazines, which contains a database of literary magazines. Even though this site has been made for writers, it looks to me as if there is some nice genre information in there, as well as other information that could appeal to the public. *If* this database would be made available in specific formats in specific ways (and if may be already), this information could be made available to the public using library tools, and to do it, catalogers would not have to copy any information at all, only the links. And if it turned out that the links could be automatically generated, e.g. by using ISSNs, there would be no need for catalogers to do anything at all.
One of the conundrums for catalogers is this: linked data is not their job. Building it is a job for programmers. Certainly, catalogers can be involved to help the programmers figure out which information should be equivalent to names, to titles, to subjects, and so on but this is only if programmers want the help. Often they prefer to do it entirely on their own.
So, when discussing linked data, instead of manually adding birth and death dates, affiliation, gender, etc. because RDA says to do so, we should be asking whether this information exists anywhere else on the web and if so, what it would take to make it usable through linked data.
James Weinheimer firstname.lastname@example.org First Thus http://blog.jweinheimer.net First Thus Facebook Page https://www.facebook.com/FirstThus Cooperative Cataloging Rules http://sites.google.com/site/opencatalogingrules/ Cataloging Matters Podcasts http://blog.jweinheimer.net/cataloging-matters-podcasts