 |
| The BibFrame Model, from http://www.loc.gov/bibframe/docs/images/bibframe.png |
An excellent recent Cambridge Librarians' Brown Bag lunch centred on the article 'LC’s Bibliographic Framework initiative and the Attractiveness of Linked data by Kevin M. Ford in Information Standards Quarterly (spring 2012, vol 24, issue 2/3)' available at http://www.niso.org/apps/group_public/download.php/9405/SP_Ford_LC_isqv24no2-3.pdf The Library of Congress (which is working on BibFrame) has further information if you are interested (http://www.loc.gov/bibframe/faqs/).
I had never heard of BibFrame, and I thought I knew what linked Data was, but I did not. I assumed it meant datasets that are embedded in articles, which has been a big topic of discussion amongst researchers recently, but halfway through the discussion it dawned on me that it was nothing to do with this at all.
World wide web experts, such as Sir Tim Berners Lee, want to work towards a semantic web: a web where information is sorted and systematic. Librarians sort information using authorised headings, rather like an efficient secretary devises a filing system with a limited number of named drawers so that if a user knows the name of the drawer they can find all the information in that topic in that one drawer. Most information relates to other information in many ways. Web experts have been trying to devise ways to describe different 'entities' so that when users add information to the web they can give it invisible code tags, in XML, so that other users can see what sort of an entity is being described and how it relates to other entities. Systematically sorted information can be much more useful than the current random mass of information that comprises most of the Web.
Librarians have already been thinking about these issues and have devised the Resource Description Framework (RDF) a language for describing relationships and entities between the literary works that they curate, but most library data is still written in a specific code interpreted only by Librarians, namely MARC code. The WWWConsortium and the Library of Congress want to change the language that library catalogues use so that it is more compatible with the way other things will be described in XML on the Semantic Web, and this language will be called BibFrame. Librarians will type names and places , or codes for names and places, into web forms in their offices, and library systems will interpret what they write into a link that links to the entity (we think this authoritative entity is called a namespace but we were a little unclear about what exactly a namespace was). So for instance, a record in a Cambridge Library for a book by Newton might be linked to a single authoritative namespace for Newton, to which all the Newton related information in the world would link. Hence we would not have to write 'Newton' into all the records, but only put in the code for Newton, and our readers would be able to see the word 'Newton' displayed and with it, depending on the filtering, links to all other Newton related information entities on the web.
Naturally this raises issues of filtering, data overload and censorship issues.
Standards are currently being drawn up and tested, but the experiments are at a very early stage.
The library system company Proquest are planning a system that only uses BibFrame and not MARC: all MARC records will be converted to BibFrame records for their future users. It is less easy, apparently, to turn BibFrame records into MARC records at the moment.
Discussion centred on the relationship between opening up library datasets to public scrutiny and the opportunities this offered for commercial entrepreneurs to interact with our records and users. The need to buy commercial services to filter and search records was also noted. Beneficial mashing was also discussed: a researcher had taken details of printers locations from a library dataset and been able to map historical printers from the data. Comparisons were drawn between the disappointing way that the Open Access movement had been taken over by commercial enterprise and it was feared that Bibframe would lead to further need to co-operate with commercial service providers.
For a discussion amongst a room of people who all freely confessed beforehand that they had never heard of any of the concepts involved before reading the article, it was quite an impressive discussion. But that is Cambridge for you.
Comments