information architecture & critical history of software (PhD research) in Toronto

view my cv

Quinn DuPont studies the critical history of software technologies, focusing on metaphysical, historical, and political issues. He has recently been studying the history of email and developing an argument about the modes of production for software development. Quinn is currently a MITACS Enhanced Accelerate PhD Fellow and iSchool PhD student in Toronto, Canada.

search
reading
  • Difference and Repetition
    Difference and Repetition
    by Gilles Deleuze
  • From Taylorism to Fordism: A Rational Madness
    From Taylorism to Fordism: A Rational Madness
    by Bernard Doray
  • Questioning Technology
    Questioning Technology
    by Andrew Feenberg
« Taxonomies, structure, and information theory | Main | Chp. 5: Ongoing review of Building Enterprise Taxonomies by Darin L. Stewart »
Thursday
Aug142008

Chp. 6: Ongoing review of Building Enterprise Taxonomies by Darin L. Stewart

As I re-read Stewart's chapter on "structure" in preparation for this ongoing review a shudder rolls down my spine on the first sentence: "Staring down a large pile of unorganized, candidate terms can be daunting." Not exactly sage advice, but grist for empathy certainly. I'm currently undergoing an enterprise taxonomy project at work and by following Stewart's wise words (which is really just standard practice), I've found myself on the good end of a day of "click, paste, click" as the intranet sight gets less interesting and my MS Excel spreadsheet gets longer. This is, of course, the entrance into the the exciting part of the project and the exciting part of the book.

Stewart starts with describing what card sorting is, its benefits, and how it occurs. This is pretty standard advice about card sorting but Stewart does a fine job of covering the essentials and offers a few surprising hints along the way. In my opinion, his best card sorting trick is to employ team cards sorts in certain situations. Stewart suggests that the discussion and debate in a team card sort encourages people to rethink their assumptions—it's almost a wisdom of the crowds approach. Stewart speaks briefly about some of the current card sorting applications currently available, although misses my new favourite OptimalSort. (And, this list is going to be horribly out–of–date very soon, but that's the nature of the printed page.)

Stewart's methodology for facet analysis and construction is spot on, if not a bit predictable. A little bit of history, some simplifying assumptions, and you've got a practical and real–world facet construction methodology. Stewart is clearly biased towards a faceted taxonomy, since he opines about its benefits in chapter three and then devotes most of chapter six on its methodology, but I thnk this is a valid approach (and since facet analysis and construction is more difficult, it makes good sense to spend some time getting the details straight). The chapter ends with a description of the traditional types of relationships found in a taxonomy, including helpful pictures for visualizing the relationships.

Technorati Tags: , , ,