[uf-discuss] hCard to represent simple entities (was: Tentativeproposal...)

jim at eatyourgreens.org.uk jim at eatyourgreens.org.uk
Fri Jan 4 03:49:45 PST 2008


I thought about this a bit more this morning, and it sort of creeps into
the territory covered by rel-tag too. For instance, the blog post I made
yesterday has astronomical names that are being used as tags - '2007 WD5',
'Mars', 'solar system', 'Shoemaker Levy 9'
(http://www.nmm.ac.uk/rog/2008/01/will_an_asteroid_hit_mars_in_j.html).
Those are all strings that could be marked up, semantically, as the names
of things.

TEI is a markup language that's mostly used in the humanities for
digitising written texts such as books, plays, poetry and letters. In terms
of what it can teach us, I guess a lot has been written by the TEI
community on how to mark up the structure of a written document eg. how to
mark up dates and times, how to mark up names of things. There's a good
overview at this URL:
http://tei.oucs.ox.ac.uk/Oxford/2007-02-13-oucs/talk-overview.xml

We used TEI-Lite years ago, to digitise an archive of papers relating to
the explorer Matthew Flinders. Specifically, we wanted to index papers
according to people, places, ships etc. mentioned in those letters. The
indexing had to be good enough to recognise that eg. 'The Revd Mr Tyler'
and 'Tyler, (Reverend) William' are semantically the same. TEI-Lite is good
for this sort of problem.

That archive is available in HTML at
http://www.nmm.ac.uk/flinders
but the HTML version loses the semantics of the underlying TEI documents,
which are hidden away in a database and, therefore, unavailable to the
serious researcher. We could open up the original semantic documents by
giving links to the XML versions, but it might be useful also to try and
preserve the TEI semantics in the HTML documents. There's a post on
Semantic Humanities about this problem, but with regard to RDF too:
http://semantichumanities.wordpress.com/2006/08/26/when-is-semantic-html-not
-important/

Jim

Original Message:
-----------------
From: Andy Mabbett andy at pigsonthewing.org.uk
Date: Fri, 4 Jan 2008 00:34:35 +0000
To: microformats-discuss at microformats.org
Subject: [uf-discuss] hCard to represent simple entities (was:
Tentativeproposal...)


In message <D0531DB4-856C-4CC9-BD56-4539D4D4AA13 at eatyourgreens.org.uk>,
Jim O'Donnell <jim at eatyourgreens.org.uk> writes

>> For clarity, the former can be distilled to:
>>
>>         hCard is for representing people, companies, organizations,
>>and
>>         places

>Reference strings, in TEI markup at least, can also refer to the  names
>of books, ships, plays, films and pretty much anything that can  be
>given a name. hCard works for people and places, but is it general
>enough to cover those cases?

I think ships are an edge-case for hCard.

For books, plays and films, I would think that's a job for a "citation"
microformat, once we have one (and one is surely needed).

[One could argue that a physical copy of a book could have an hCard,
with an extended-address of "Shelf 54, Floor 3, Anytown Library"; but
that's really stretching the logic.]

As for "pretty much anything", I'll leave that for others to decide ;-)



I'm not familiar with TEI:

        "a consortium which collectively develops and maintains a
        standard for the representation of texts in digital form. Its
        chief deliverable is a set of Guidelines which specify encoding
        methods for machine-readable texts, chiefly in the humanities,
        social sciences and linguistics.

        <http://www.tei-c.org/index.xml>


- what does it have to teach us?

-- 
Andy Mabbett
_______________________________________________
microformats-discuss mailing list
microformats-discuss at microformats.org
http://microformats.org/mailman/listinfo/microformats-discuss


--------------------------------------------------------------------
mail2web.com - Microsoft® Exchange solutions from a leading provider -
http://link.mail2web.com/Business/Exchange





More information about the microformats-discuss mailing list