I think the opposite about the ontologies actually - the whole point
is that you can mix and match ontologies at will, each one
specialised for describing a particular kind of 'thing' (people,
documents, publications, etc). However, I agree that the bibtex
ontology is not much use for anything other than describing the type
of document, so it seems this description (document type) would be a
good addition to PRISM, if it isn't there already. That would leave
you with just three ontologies (DC, FOAF and PRISM).
Does that mean Google's using the scraped RDF for their indexing?
alf.
On 14 Sep 2005, at 19:12, Matthew Cockerill wrote:
> Google Scholar felt very strongly about avoiding the need for an
> extra round trip to get the RDF. Anything which doubles up the
> number of HTTP requests isn't good when you're trying to manage
> tens of millions of documents and keeping them up to date.
>
> And embedding the RDF as a comment is already in very common use
> for CC metadata - so shouldn't Piggy Bank really support the
> identification of islands of RDF in comments like this?
>
> Meanwhile though, we will try at some point to add in a proper link
> to the RDF as a separate URL as well, sure....
>
> Re: the use of bibtex:article - it seems frustrating to have to
> pull in *so* many different ontologies to express bibliographic
> information.
> The bibtex ontology overlaps with the PRISM ontology quite a lot,
> no? So pulling it in seems in some sense messy - but maybe I'm not
> looking at this in the right way.
>
>
> But it seems to me that it would really help to have an ontology
> that pulled together what is needed to serve the practical needs of
> publishers than the messy ad hoc combinations of Bibtex/FOAF/RSS/DC/
> PRISM (but which could of course express equivalences with those
> ontologies)
>
> Analagous to the National Library of Medicine's publisher DTD.
> dtd.nlm.nih.gov
> Maybe the NLM can be persuaded to create (or endorse) an
> associated bibliographic ontology?
>
> Matt
>
>
>
>> -----Original Message-----
>> From: Alf Eaton [mailto:lists_at_hubmed.org]
>> Sent: 14 September 2005 17:44
>> To: general_at_simile.mit.edu
>> Subject: Re: structured bibliographic info for BioMed Central
>> articles
>> now available as RDF
>>
>>
>> On 14 Sep 2005, at 18:30, Leigh Dodds wrote:
>>
>>
>>> Why not also make the RDF data separately linkable,
>>>
>> avoiding the need
>>
>>> for other tools to scrape the data out of the HTML? For
>>>
>> IngentaConnect
>>
>>> I was going to add an "autodiscovery" link style link to associate
>>> the RDF representation with the HTML version
>>>
>>
>> I agree. In particular, Piggy Bank isn't able to automatically
>> recognise the RDF embedded in the page and make use of it.
>>
>> Also, is there a reason why it's an RSS item, rather than something
>> from a more appropriate ontology (say, a bibtex:article, for
>> example)?
>>
>> alf.
>>
>>
>>
> This email has been scanned by Postini.
> For more information please visit http://www.postini.com
>
>
Received on Wed Sep 14 2005 - 18:37:34 EDT