Museums and the Web

An annual conference exploring the social, cultural, design, technological, economic, and organizational issues of culture, science and heritage on-line.

You are herejtrant's blog

jennifer trant's Blog

Social Tagging and Access to Collections: podcast + some followup thoughts

can we learn from tagging?: Tagging might give us some insight into visitor interests...i was supposed to be in Brazil this morning, talking about "Social Tagging and Access to Collections". unfortunately, visa problems made that trip impossible. the presentation was made via podcast; i've put the files on-line.

it's an introduction to the issues that we're exploring in the project – how social tagging might enable access to art collections – focusing on the gap between user interests (as we know them from queries and reference questions) and museum documentation as created by and for professionals. This work is more formally presented in other steve papers (see below).

that gap exists partially because our documentation standards are theoretical; designed by and for professionals to serve museum functions. i've participated in many initiatives that explored aspects of access to collections – CHIN, FDA, AITF/CDWA, CIDOC, MESL, AMICO – winner take all in the aconymble contest – and don't dispute this as a valid first premise. it's a place we had to start from, but it may not put us in a good position to support public access.

museum data is still seldom shared beyond the bounds of local systems; and when it is, in early cases like MESL or building The AMICO Library, we discover that institutions adapt standards, rather than adopting them. merging data from disparate sources is hard. we compromise on what has to be merged and end up with thin (lowest common denominator) descriptions that may not be adequate for differentiation between similar objects, and may not support user needs.

i'm more convinced, since the discussion with seb and mike about the contribution of tags, that we need to look at the types of terms we're getting as tags and the types of terms searched, when we build on my preliminary analysis of Guggenheim search logs as part of the term analysis.

that way, we'll be able to learn a bit more about the relationship between tag, searches, and documentation structures, and do it within the context of searchers' needs.

user contributed content @ the Whitney

making your markThe New York Times has published some great photographs of artist Rudolf Stingel's installation at the Whitney Museum of American Art. to accompany a brief piece by Roberta Smith.

With the gallery walls themselves as canvas, visitors have been asked to complete the work done at the Museum of Contemporary Art in Chicago, whose work comprises the upper level of the piece.

The result is a definite contrast to the white cube.

/jt paper for ichim07

steve tagger interface: show tags: The steve tagger with the 'show tags' option, where the user sees tags already assigned to a work.

we've just released the paper that provides background to the demonstration Susan Chun will be giving at ICHIM07. it begins to present the methods for's IMLS-funded national leadership grant (research), focussing on our data collection strategy and the steve tagger.

it also includes some preliminary results about attitudes to tagging amongst museum staff, and early analyses of tags collected. Early results in prototype tests are holding up. We're seeing a greater than 75% new tag rate:

"Of the tags assigned to all works during Term Set 1 (March 27– July 11, 2007), 76.5% (7,973 of 10,418) were not found in museum documentation." (Trant 2007)

That means that, during our first phase of data collection, more than three-quarters of the tags assigned in the steve tagger were not found in the museum's own documentation for those works.



Trant (2007). Trant, J., et al., The eye of the beholder: and social tagging of museum collections

, in International Cultural Heritage Informatics Meeting (ICHIM07): Proceedings, J. Trant and D. Bearman (eds). Toronto: Archives & Museum Informatics. 2007. Published September 30, 2007 at

20 countries to be represented at ICHIM07

ICHIM small logo: six-fingered handTo date, 20 countries will be represented at ICHIM07 in Toronto, October 24-26, 2007.

  • Australia
  • Austria
  • Belgium
  • Canada
  • Denmark
  • Estonia
  • France
  • Germany
  • Greece
  • India
  • Ireland
  • Italy
  • Japan
  • New Zealand
  • Norway
  • Scotland
  • Sweden
  • The Netherlands
  • United Kingdom
  • USA

As can be seen from the papers on-line, this ICHIM is going to be a great opportunity for interchange, and a chance to see a broad range of diverse cultural heritage informatics applications. With time still to register, i'm sure we'll add to an already impressively diverse group of delegates.


International Cultural Heritage Informatics Meeting (ICHIM07) Papers On-line

ICHIM07 - cultural heritage informatics in torontopapers are now on-line for ICHIM07 in Toronto, Ontario, Canada, October 24-26, 2007. over 50 groups will be reporting at ICHIM07 on the state of the art in cultural heritage informatics, and plotting the issues we face in the future. see the list at

MW2008: the proposals are coming, the proposals are coming ...

MW2008 logo

well, true to form, we've been watching the proposals for Museums and the Web 2008 flow in all day. what is it about our always on culture that's embedded this last minute madness in the way we work?

i know that even though it's late on friday (most places) we'll see some more filter in tonight. and then there's tomorrow and the next day (did anyone notice it was a weekend?); we didn't when we set the deadline...

ah well – we know that there's a problem in museum culture: a lack of time to be reflective, even though that's one of the things we try to provoke in our work [thinking, learning, seeing anew – aren't these what we want our visitors to do?].

as our small contribution to collective memory, we've insisted on written papers for Museums and the Web (and we keep them all online).

so now, make that reflexive jump towards the next deadline: September 30, 2007. Museums and the Web, Montreal, April 9-12, 2008. Call for Particiption. we want to see your best work, and participate in the process of discussing, documenting and sharing it. we can all take some time out to learn.



Feeling over-connected? try NOSO

NOSO Header

The NOSO site has been out there for a bit -- indeed there was a presentation @ the de Young the end of June -- but i don't recall anyone mentioning it in the museum sphere.

it's a great observation on our dependence on tools to organize our selves, and our [social] lives, and our need to maintain identity through personal space.

drop in, tune out ... i think this is why we have an island.


classification, folksonomy and lyrics

album cover


got a corvair in my yard

hasn't run in 15 years

it's a home for the birds now

it's no longer a car.


Jim White, Corvair, No Such Place

yes! new mw2007 photos

Jon Pratty's posted some new photographs of Museums and the Web 2007 in our MW2007 flickr group. it's nice to know that i'm not the only one who takes a while to get my pictures organized -- and to see some changes in the images in the right nav bar of this site. no matter how random flickr says it is, it definitely favours some images over others (or seems to start in the same place). maybe there's a setting i've missed?

thanks Jon!


on research in the museum context - sparked by Seb Chan's post on the update

thanks seb, for the rather nice write-up of the update podcast.

you've hit on one of the real problems of conducting research in the museum context -- the task of making what seems 'academic' relevant and useful. it's hard, in our project-driven culture, to step back from what we think we 'know' to see if there is any evidence for our prejudices. Lorcan Dempsey blogged about this not long ago, pleading for an 'evidence base' in library catalogue development.

we need research in order to do what we do better, but we can't have research for research's sake in a sector devoted to public service.

i've been thinking about this in the context of how we structure and report both the research agenda and its results. obviously, there's a need to distill the research rationale into something that's easily digestible to the professional. but doing that without dumbing down is a challenge. make it seem to easy and the need for the research gets lost; make it too applied, and the product takes the place of the questions and the research environment starts to look like an answer in itself.

Xavier @ ICHIM03 in Paris

Xavier @ ICHIM03 in Paris: Xavier and the sixth finger, at l'École du Louvre./jt

Xavier @ ICHIM03 in Paris: Xavier and the sixth finger, at l'École du Louvre.

/jt update for the Dublin Core Annual Meeting

I've been invited to give an update on the activities of at the Dublin Core Annual General meeting. While I can't be in Singapore, this invitation did provide the motivation to do a podcast providing an update on where we are with the research and highlighting some very early research results. You can download the podcast here.

We're definitely seeing differences in behaviour between tagging on the single-institution site at The Metropolitan Museum of Art and the multi-institution site at And we're seeing differences in tagging depending on whether users see object descriptions or not, and whether they see works of art in related sets or not.

The current data collection phase is exploring the effect of showing others tags. In the fall, we'll move on to explore what happens when users can select works to tag. We had some great discussion about what 'choice' means at the steve summer meeting on Grindstone.

Tagging game at the McCord Museum, Montreal

The McCord Museum has launched a 'keyword matching game'. Players are asked to supply up to 10 keywords for each of a group of images, within 3 minutes.

It took me a few minutes to figure out the interface -- and i found myself peering at the images [maybe that's just because i've had a long day at the screen, but i was looking for details, and found i was squinting.]

McCord Museum of Canadian History: Scoring Screen: The scoring screen from the McCord's keyword matching game. term set 2 data collection launched

we've just deployed the second phase of the tagging experiment. at

the steve tagger (a piece of open-source software) is a key tool in our IMLS-funded study of the contribution social tagging and folksonomy can make to on-line access to art collections. throughout our experiement we'll be varying the interface of the tagger to find out what encourages

and we'll also be studying the results of tags, to see if they are:

  • real words (we're using word net)
  • terms from our discipline (we're using the AAT and ULAN)
  • new to the museum (we're comparing to museum documentation)
  • appropriate to the work of art (we're doing term-by-term review).

the research process is described in the attached diagram.

we're looking forward to sharing the results of our study with the community. If you'd like to participate, please come by.


ICHIM07: Preliminary Program Announced


ICHIM07 - International Cultural Heritage Informatics Meetings
Toronto, October 24-26, 2007

The preliminary program for ICHIM07 is now on-line. Thanks once again goes to the Program Committee who helped review proposals and recommend selections. Speakers are coming from 20 countries around the world: