[openbiblio-dev] Best practice for name format in BibJSON

Sat Feb 25 09:18:55 UTC 2012

On Fri, Feb 24, 2012 at 1:35 PM, Jakob Voss <jakob.voss at gbv.de> wrote:

> Hi,
>
> I had a look at http://bibjson.org/ and agree that it has some benefits
> compared to BibTeX and RDF-based formats - there won't be one and only
> format anyway, but different forms for different use cases.
>
> I am glad you find it a useful approach.

> Apart from the missing mapping to Citation Style Language JSON [1] I only
> found the following issue problematic: there is no best practice guideline
> how to format object names. Some example snippets from http://bibjson.org/
> :
>
>

> "author":[
>        {"name": "Richard Jones"},
>        {"name": "Mark MacGillivray"},
>
> "journal":{
>    "name": "A really great journal",
>
>
> [... other name metadat snipped ...]

I'm speaking for myself, but the Openbiblio2 project (which currently
supports the software, architecture and example development) does not see
itself as creating canonical representations. Rather it creates an
infrastructure which can support a variety of vocabularies and practices.

Names are hard, often very hard, and many authorities are wrestling with
them. BibJSON can represent these different approaches because it has
support for namespaces and syntactic flexibility.

BibJSON does not attempt to normalize, disambiguate or canonicalize. Rather
it provides a powerful, hopefully simple, approach where the information
can be clearly represented, stored and processed. We recently had a
discussion on the OKF open-access list as to whether the authors of the
Panton Principles should be cited with first names. This gave rise to
several emails. I can guarantee that there are thousands of similar
approaches out there. So, if the authors wish to hold first names, we find
an authority which supports the concept.

I am not keen on implicit semantics. "author: Henry James" and author:
"Henry, James" is a very fragile way of managing info. So, assuming *I*
care, I have to find a vocabularly that supports this. If there is one out
there I look to see if it has a namespace and use it. If there isn't I then
I have to create one and give it a namespace.

I am expecting Wikipedia and other public-facing organizations to tackle
this problem and, by default, I would us a Wikipedia vocabulary (with the
appropriate namespace).

-- 
Peter Murray-Rust
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
+44-1223-763069
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/openbiblio-dev/attachments/20120225/75671cba/attachment.html>