Getting comments on this site from Tallinn

If you’re at the Tallinn conference and you haven’t got a WordPress account password, just email me directly and I’ll put your comment online myself.

Email me at editor@24hourmuseum.org.uk

Jon Pratty

Collections tags / the Power House

Thought you might be interested in a different view of the Power House collections website. In the V&A meeting we took a baffled? or just not convinced? look at the user generated tags to its collections database. Students (c. 20)  in my Museums & Digitisation course reviewed the site as 1 of 8 for authority, usability, interest, other characteristics. The PH came top with 19/20 votes for “I think this site is used a lot”. They liked the tags (no prompting!) not because they offered a structured way into the collections, but because they generated curiosity and interest – what is it that that other person is interested in – what can that peculiar thing be? It’s like ‘other people are searching for …’ so the function of the tags is completely different, they truly socialise the collection. Another well known example of folksonomy (must be loads now) is the San Francisco Art Museum where their docents produced new categorization terms, a ‘word soup’. But these are to make for better indexing and searching, so they are essentially towards a better categorization system, not truly social software.

Thanks for an enjoyable and stimulating meeting Frances,

Suzanne

Dublin Core 2007

The call for papers for DC2007 (http://dublincore.org/workshops/dc2007/cfp/) shows concern with lots of the stuff that’s preoccupied us at the workshops so far, particularly the mapping of knowledge domains through formal SW means and how social tagging can fit in.

Jeremy

Semantic community tagging for ‘power users’

One thing which concerned us for some time in Newcastle was the ‘interface’ or possible relationships for users and staff between contextualities built by folksonomy and those more formal structures built by domain experts such as curators.  We’re not at the answers stage yet I think, apart from an agreement that both approaches are valuable and can complement each other. Maybe its simplistic but I like to see the ‘formal’ side acting as semantic glue sticking folksonomies together.

Things are moving along elsewhere. DBin looks interesting: http://dbin.org/ RDF semantic tagging of discussions by domain experts. Danny Ayres (Raw) has been blogging about SWAP 2006 in Pisa and this among other things, raised his interest.

Mike

3D vs 2D and the richness of folksonomies

In Brighton I briefly stumbled through an attempted explanation of something I think may be worth expanding on. The context was discussion of the relative values of curatorial and folksonomic terms, and what I wanted to get at was (a) terms for describing what you’re looking at may not be the same as you would use if searching for the same thing, in part because (b) what you describe when you’re looking at something varies according to how you see it: whether it’s in front of you, 3D and full size (or not), or on a screen in 2D (and most likely shrunk). The 2D/3D thing was explored in a presentation at MCN by Elise Lewis and I’ll snip my notes from the conference here (so please excuse the note-like nature):

Lewis (U
North Texas) – user interactions with 3D objects on the web

No common definition of “3D image”; lots of viewers, functionalities and needs. Lots of research in e.g. games and commerce but not in cultural heritage. Did a pilot study with savvy grad students. Their descriptions of 3D images from Arius* were mapped to hierarchical levels of perception:

• colour
• shape
• texture
• object – name, generic description
• action
• location – in this case used to describe the location on the object (view). In 2D, a geographical term
• affect

Descriptions also on 2D versions of same item. More description on 3D, especially action, object, colour and location. Occasionally more terms used for texture in 2D that 3D but descriptions are much richer in 3D as more facets emerge. Need a survey to find out how and why these images are being used.

As such objects become more common metadata and controlled vocabulary issues (e.g. for location) will become more urgent. Will user generated content be helpful or not? There will be management issues too – MD, systems, retrieval.

* Arius is a company who work with Royal Ontario Museum, they’ve done some cool colour laser scanning stuff for them

Curators inevitably get to see stuff in 3D, and quite possibly bigger, too, so it follows from the above that quite aside from being more “academic”, their descriptions may also be much richer simply because of their interface with the item in meatspace. Which is not to say that they should always be present, let alone privileged above other terms used for an item.

Jeremy O