In terms of Task Analysis, one could make the argument that sure users may be initially exploring but ultimately its all about completing a task(eg, research). That being said, you’re right in that TA is more relevant to software dev vs reference site such as an online encyclopedia. But as you also may be aware, more companies are follwing the web application model of software development and lines between exploration and pure task based functions are becoming more and more blurrd.

So the question is not either or which is better or worse for either situation, but as you alluded to, a more solid approach and sensitivity to how people consume information and differing ways of doing so.

You seem to be talking about the information retrival community, but there is a part of the Information organization community that is exploring on the area of seeking unknown information through a culmination of organization methods. Enabling Serendipitous discovery is the word I guess.

Although this community is exploring that, I have realized its hard validate the effectiveness of organization as there is no proper testing mechanism available. Thus traditional usability methods are used, which in my opinion doesn’t really give “real” results.

Melvin Kumar

Very interesting musings. As an ex-LIS person turned IA and now doing more interaction design and usability, I’ve struggled with pretty much the same issue you put forth above.

I agree the the usability field hasn’t come to grips with exploratory information seeking. The HCI community also tends to misunderstand information seeking to some degree as well. Solving complex information problems and creating meaning from the information we find can’t always be accurately described by a task analysis. There’s an interplay of cognitive, social, cultural, orgainzational, affective and linguistics factors to consider.

But LIS folk do indeed recognize that there are users at the other end of their content - at least in theory. For instance, the notion of human-information interaction already exists as a sub-field within information science. The iSchool in Washington even has a center for it:
http://www.ischool.washington.edu/chii/index.html
Other schools refer to such concepts as “human information behavior” or the like. What’s lacking from the LIS field - in my opinion - are the practical solutions to keep up with a quick-paced business world that IAs create on a daily basis. It’s really the typical gap between theory and practice found in just about any discipline. Morville and Rosenfeld have done great work “pulling” from LIS, but there is little “push” as far as I can tell.

I’d be happy to point to you a butt load of information behavior research if you want. Just let me know.

Cheers,
Jim

Donna Maurer has an insightful post comparing the task-orientation of HCI to the iterative information seeking behaviors we see on the web. Trying to apply