Re: Just Starting Out

From: Jason Foster <retsofaj_at_mac.com>
Date: Mon, 15 Aug 2005 08:28:51 -0400

> I've got the next step down slightly differently - finding a way of
> expressing the data in RDF in a useful fashion (i.e. one in which you
> capture as much of the info as possible). Essentially find appropriate
> vocabularies/RDF schemas. Then look at mapping/translating the data
> from its current form to an RDF form. With databases etc the ideal is
> a continuous/two-way mapping, but in practice at this point in time an
> export dump and one-off translation may be the best approach
> available.

This is where I very much get torn. On the one hand, I have a number
of tools that are "optimized" to handle particular types of
information but that need aggregation. On the other hand I like the
idea of not having to worry about exporting, synchronizing, etc.

To be honest, this feels a lot like the issues I have with Apple's
Spotlight. Spotlight has translators (analogous to the RDF mapping/
importing scripts) and a common search interface. I'm still not
comfortable with Spotlight, and I think it's because I just don't
trust it to be current, properly indexed, etc. I'll concede though
that this is more a psychological issue that a technological one.

> I must confess I've not looked at Haystack since its (slightly
> disappointing) early versions, it may now offer what you're after
> (time I started a download, methinks).

I played with it a little yesterday. It certainly looks interesting,
but I think there will need to be much more playing before I know if
it makes sense as a repository.

> An option for dealing with loose text notes might be to paste them
> into a blogging system, export as RSS 1.0 . WordPress with the FOAF
> Output plugin can provide fairly rich metadata, there's the category
> stuff and links within content are pulled out as dcterms:references
> statements (e.g. as at [1]).

Very cool idea; thanks!

> I'd be tempted to experiment with text file based storage (i.e.
> RDF/XML), at least initially. I found it straightforward to set up
> with my own data, (see [2]) . I've no idea of the current
> Longwell/DSpace status. At some point, personally I want to use
> Longwell/PiggyBank alongside a Redland RDF store, but that's another
> story.

Ick! Call me lazy, but the idea of having to maintain a set of XML
or N3 files manually gives me hives. I'm definitely in the hunt for
a tool that provides a nice, intuitive, lazy-person-friendly editing
interface.

Jason
Received on Mon Aug 15 2005 - 12:25:07 EDT