Jennifer Crowe
Since 2002
Works in Brooklyn United States of America

Discussions (1) Opportunities (0) Events (0) Jobs (0)

Re: Re: Metadata

Hi All,

I just got back from a conference in Leicester, England (The Museum - A World Forum) where I presented a digital artwork that they, along with Phoenix Arts (local contemporary art organization,) commissioned for the conference at a local museum. I was also asked to talk about taxonomy and archival issues for work such as mine in the context of museum collections.

Of course, the Artbase and the Variable Media Initiative came up.

Since I was very involved with the early work on the Artbase, I'm happy to see the conversation about metadata picked up again on Rhizome.

My responses to Rick's original questions are below:

>1) Do Rhizome's vocabularies need to be compatible with other metadata >standards? If so, which, and how much?


A willingness to be compatible with existing standards whilst building unique terms for artworks of this sort is crucial to accessibility and, therefore, the future viability of the Artbase project.

>2) What can we propose here that Rhizome can practically accomplish >given limited resources?

A hybridized, tiered system where existing standards are the starting point and rhizome folksonomies create the bulk of the lower level descriptive terms will probably be the best model moving forward.

>3) Currently the metadata that uses vocabularies is divided into type, >genre, and keywords -- are these categories sufficient? Should we add >others?

Possibly, but I would use these existing well-established metadata categories as a starting point and then expand from there as needed. Most likely, you'll have to work from the bottom up here in order to gain any legitimacy.

More detailed analysis of various organizations metadata systems is still needed. I'm not at all saying do what everyone else does, but there must be an informed starting point.

3) Currently the metadata that uses vocabularies is divided into type, genre, and keywords -- are these categories sufficient? Should we add others?

Possibly, but I would use these existing well established metadata categories as a starting point and then expand from there as needed.

>4) Do we want to enhance/ elaborate/ add on to our existing descriptive >terms or keep the current controlled vocab as is, and make folksonomy >also an option?

Do both.

Folksonomies are works in progress. The results are tools by which consensus emerges. Any "consensus-es" reached will only be relevant if there are mechanisms to play nice with, but not be a slave to codified standards.

I'm not afraid of standards and I don't think anyone here should be. They are merely tools that can be manipulated if you are just diligent enough. Until the larger library science/informatics people come around to folksonomies as a mainstream way to drive standards, we'll have to just bear with them and use a flying under the radar approach to make change. It is happening already, albeit rather slowly and that's probably OK.

However, if folksonomies govern the formation of all Rhizome standards without an examination and strategic, selective adaptation of existing well-established standards the Artbase will become further ghettoized.

>5) who is the artbase for? Who is its audience, and how does that >affect our re-design of the metadata.

That is an excellent question.

Collections have different needs dependent upon context and purpose. The current context and mission of the Artbase seems rather fuzzy. Since I was involved with it back in its earliest days, it has evolved from an archive-in-progress/development with an archival/preservation mission, to primarily a list of informative links out to artworks, described with Rhizome-defined keywords.

How do the needs of a collection of art objects (like adaweb, Artbase's cloned objects) differ from a list of links (the current majority of the Artbase)? Rhizome must address this question for the Artbase before proceeding.

The Artbase first needs to decide if it's an archive of art objects or if it's an online resource that links out to art, as this will surely inform the kind of structure is needed, as well as the terms that come out of any folksonomy exercise.

Tag clouds, collective filtering, etc. are the easy part. It's taking the results to the outside world as part of an accessible metadata structure and accompanying documentation that will be more difficult and a longer-term exercise.

I would also argue that in additional artists, curators, art fans, etc., a key audience should be the museum/library informatics community. These are the people who have traditionally set the standards and who still currently hold the power.

I'm not saying pander to them. Rather, show them how to do it better.


PS: On an added note, Pall's suggestion to archive code is excellent. Rhizome must first be prepared to clarify the Artbase's mandate regarding preservtion/archiving, however, in order proceed effectively.

Richard Rinehart wrote:

> Thanks Sal,
> You summed it up nicely. It's great to see convergence on some of
> these issues on the list.
> Richard Rinehart
> At 12:44 PM -0400 4/26/06, Sal Randolph wrote:
> >I second rob & david's arguments about why folksonomies are great,
> >and I think they would mix amazingly well with rhizome membership
> >(all members would get to tag the artbase as they like). tag cloud
> >of the artbase indeed!! It would naturally evolve as the field of
> >inquiry evolves. Also as someone who has implemented a tagging
> >system with freetag recently, it's *really* easy to do (freetag is
> >php, and I know rhizome's using ruby, but it doesn't look all that
> >hard to write one either, just from surveying the code).
> >
> >Personally I would advocate for a double system, as some have done:
> >free folksonomy tagging by everyone, and then a layer of curated
> >language for either genre or keyword. multiple points of
> >intellectual access are a good thing.
> >
> >
> >On Apr 26, 2006, at 7:51 AM, wrote:
> >
> >>Quoting Richard Rinehart <>:
> >>
> >>>I'm curious about the statement you made below Rob, that any
> >>>folksonomy can be made compatible with standards using a good
> >>>thesaurus. Do you have an example of this?
> >>
> >>I don't have an example I'm afraid. It's more a strategy I had in
> mind for
> >>paintr ( Folksonomies and taxonomies
> are both
> >>formalisations of human language, so if my RDF doesn't contain the
> >>word "blue"
> >>but it does contain the word "color" I can locate my tag in the RDF
> using
> >>wordnet or a thesaurus.
> >>
> >>>Your note on the AAT is very (VERY) well taken. Yes, the AAT is
> >>>not yet a good resource for terms for new media art, yet it is the
> >>>single standard used most by museums and other organizations
> >>>collecting new media art. So, one strategy would be to ignore the
> >>>AAT as irrelevant; but another might be to work with the Getty to
> >>>update and improve the AAT with relevant terms so that (digital)
> >>>community-specific practice becomes (museum) community specific
> >>>practice rather than creating a ghetto (though I'm not sure which
> >>>is the ghetto of the other here :) In the past, the Getty unit
> >>>that had maintained the AAT had expressed interest in updating the
> >>>AAT based on feedback from the relevant community (us).
> >>
> >>Yes I think that might be a very good project.
> >>
> >>Possibly collaborating to make AAT aware and having a
> process to add
> >>more terms relatively quickly as they come up? So in artbase have a
> list of
> >>terms you can choose followed by an "other" checkbox that people
> could add
> >>terms they felt weren't in the taxonomy. We (the Rhizome community)
> >>could then
> >>keep an eye on those and see if they should go into AAT.
> >>
> >>A folksonomy might be more democratic & easier to implement though.
> :-)
> >>
> >>
> >>On the subject of proprietary software it might be an idea for
> Rhizome to get
> >>licenses for Windows, ASP, IIS and so on so that software
> >>unfortunately written
> >>for them can still be run in the future. In a few years time having
> >>this stuff
> >>available for galleries to hire might actually provide a revenue
> stream. ;-)
> >>
> >>- Rob.
> >>
> >>+
> >>-> post:
> >>-> questions:
> >>-> subscribe/unsubscribe:
> >>-> give:
> >>+
> >>Subscribers to Rhizome are subject to the terms set out in the
> >>Membership Agreement available online at
> >
> >+
> >-> post:
> >-> questions:
> >-> subscribe/unsubscribe:
> >-> give:
> >+
> >Subscribers to Rhizome are subject to the terms set out in the
> >Membership Agreement available online at
> --
> Richard Rinehart
> ---------------
> Director of Digital Media
> Berkeley Art Museum/Pacific Film Archive
> ---------------
> University of California, Berkeley
> ---------------
> 2625 Durant Ave.
> Berkeley, CA, 94720-2250
> ph.510.642.5240
> fx.510.642.5269