Andy Edmonds: Academia

Fri, 20 Dec 2002 14:46:14 GMT

Work on Uzilla.org is picking up -- specifically the "A Day in the Life" study. I had hoped that the creative commons would offer an appropriate license for the resulting data set, but Attribution-NoDerivs License.

However, how do statistical analyses fit into the notion of derivative works? And what if someone discovers something interesting through, say a markov chain model, wouldn't we want that model to be made available computationally? On the other hand, maybe I don't want a user experience consultancy to claim the data set as a propietary for-resale resource.

Tue, 10 Dec 2002 17:36:45 GMT

My research team needs a large corpus of 5 letter words. WordNet doesn't seem to have a public length based interface, but I can just import to mySQL.

Tue, 10 Dec 2002 06:46:14 GMT

I don't know whether to be dismayed or enthralled at the prospect of an rich media ad-oriented delivery platform on tablets. On one hand, a dedicated presentation layer tends to improve usability -- on the other hand, this kind of power play in client software tends to steal aspects of user control.

Sat, 30 Nov 2002 17:49:02 GMT

I'm slowly stamping out my 1998 design in the surf*mind*web collection. Lots of good indexing this morn too, check out the new links page. This is what drives the lower right link list.

Sat, 30 Nov 2002 16:55:24 GMT

Interesting, a php/mysql combo specifically designed for the task of research. D'Lib's latest issue contains a summary, Software for Building a Full-Featured Discipline-Based Web Portal

Pre-Blogging

Thu, 28 Nov 2002 00:20:29 GMT

A cognitive science student's pre-blog: monicsoft. My "pre-blog" was born in 1996 and crafted with frontier -- to auto build the navigation. There's something to be said for a strong topical organization to a research oriented weblog, but the synergies in keeping up with the blogdom are also nifty.

Monicsoft seems to be working on a nice abstraction layer on a hypertext authoring system with rich support for levels of granularity on conceptual chunks -- expressed as tooltips. I guess, you could make a case for ditching the episodic rendering and just providing recent changes in RSS.

Mathius M. has another interesting approach to his thesis communication, in keeping with the hypertext topic. Mathius is publishing correspondence and maintaining an extensive web of related topics and sites. Probably the best thesis-oriented site I've visited to date (he just needs some RSS!).

Castellan Award

Sun, 24 Nov 2002 14:03:10 GMT

Just back from a trip to Kansas City for the Soc. for Computers in Psychology/Psychonomics. My paper on Uzilla got the "Castellan" best student paper award. It was a fun trip, but I'm glad to be out of the low level experimental psychology game and into more applied work. Things like the neighborhood effects of subordinate homographs in priming with truple based presentation just doesn't interest me the way it did 10 years ago.

Wed, 13 Nov 2002 02:47:57 GMT

Micah Alpern has published some good tips on introducing blogging into a workgroup. The info is on Radio Community Server, but I'm looking to Pivot, a php dsn-less solution, as a lower cost starting point.

Thu, 31 Oct 2002 14:05:32 GMT

Pivot, a php blogging tool, looks to be pretty robust - and lightweight. It does rss and doesn't require a db. Testing it out at my uni webspace. [via fragment.nl]

Sat, 26 Oct 2002 22:46:05 GMT

Information systems research: towards irrelevance?.

Here's an interesting Seb's Open Research:

paper on the severe state of disconnect between IS academics and practitioners (pdf). Thus Glass (1997) is led to comment that "the academic picture of the industrial world (and vice versa) is both skewed and disdainful", while Pike(2000) remarks that "we see a thriving software industry that largely ignores research, and a research community that writes papers rather than software". [...] After all, didn't science itself get started as a hobby?

One of my missions in grad school is to write *lots* of software. This is more important as a usability guy than it would be as a computer scientist even -- because it's all about observing actual use and changing, significantly, the way software operates. Veritable paradigm shifts in functionality... a fully realized training wheels ui, pie menus, "velocity sensitive UIs", adaptive browsing, etc.

I do find that the scientists producing software in my field are often successful. However, too often, the systems developed are limited prototypes, hand crafted for just one purpose. Working within the Mozilla codebase solves this problem -- creations are deployable to every major platform and a strong open source spirit exists. This allows a good idea, even if poorly implemented initially, to pick up a life of it's own.

Sat, 26 Oct 2002 04:19:46 GMT

Some very interesting work on dynamic, usage based sitemaps. This is an especially important link if you've never checked out Citeseer from nj.nec.com. With cross linking of citations and opt-in submission processes, this is the future of research.

Semantically Rich Markup

Fri, 25 Oct 2002 03:23:08 GMT

Tantek's blog is a welcome addition to my awareness. The call for Semantically Rich Markup is a welcome one. I'm not hurrying to css, but am rather fascinated with the concept if not the buzz?

Are there any tools for semantically rich markup? The whole web would be an outliner if the heading tags (h1, h2, etc) were containers. There not, so web text isn't a hierarchy -- but conversations on the web often are.

Some Arm chair cognitive psychology recently suggested that minds are hierarchic, but the buzz on that is more interesting than the supposition.

If I enrich my blog with semantic markup (like I do, in a kind of random fashion) what tools might understand it? Or at least make it viewable to browsers.

Tantek's call to "move west young men" is heard. Go west. The possibilites are endless.

The energy behind RDF/RSS/XFML is appreciated and useful, but there's still potential for basic HTML markup to be enriched. W3C DOM compliant browsers are fully capable of reading arbitrary attributes from markup. In many cases, XML is overkill.

Thu, 24 Oct 2002 20:43:24 GMT

I guess it's about time to start working with bib-tex.

Thu, 17 Oct 2002 23:31:42 GMT

The R-Project is a high powered programming system for statistics/graphing. Not exactly SPSS, but it looks to be a handy addition to an experimentalist's toolbox.

Thu, 17 Oct 2002 15:41:48 GMT

Fresh from the referrer logs, Peter Higgin's bookmarks are a treasure trove of CogPsy/HCI conferences, tools, etc.

Fri, 04 Oct 2002 04:21:48 GMT

Seb on Personal knowledge publishing and its uses in research.

Thu, 12 Sep 2002 19:08:35 GMT

Indeed, wireless internet on campus makes college a whole new game.

Sun, 08 Sep 2002 15:44:40 GMT

ACT-R, a cognitive modeling system, is evolving to be a useful tool for UI engineers out of it's academic roots. One of the major roadblocks, a basis in Lisp, is being addressed with a sourceforge based java port.

Read more on cognitive modeling from the surfmind collection or get a primer on ACT-R from course notes on cogmod at CMU.

Fri, 06 Sep 2002 06:35:01 GMT

Yummy... freshly crawled .gov, 1 million strong: Trec-2002:

This Year's Aims
1. To commence experiments on a recently gathered, commercially interesting, real Web crawl: .GOV.
2. To investigate methods for finding key resources in a particular topic area: Topic distillation.
3. To investigate methods for finding a particular page which has been named by the user: Named page finding.

Thu, 05 Sep 2002 22:26:41 GMT

Oddly, a google on right hand navigation turned up this rather wacky, but seemingly legit, bibliography on APA style for citing web resources

Fri, 30 Aug 2002 04:24:31 GMT

Here's one deserving significant further study... WEBKDD'2001 , spanning Predicting User Accesses, Recommender Systems and Acess Modeling, and Acquiring and Modelling Data and Patterns.

Mon, 26 Aug 2002 04:30:04 GMT

Wow... the web has saturated collegiate library systems impressively. It's an amazing boon to have access to PsychInfo, a database of journal abstracts, from a web browser, even if it requires IE to use.

Similar warts are to be found across the board, with NetLibrary, a full text book offering, having a extremely intrusive lisencing scheme and apparently, a requirement that you only use one browser window to search and view offerings combined that with extremely bad task-goal matching in the UI. For example, you can add a book to your favorites, but only before you view it. Bah humbug.

Then, things like the ACM library, are available only on campus. Still, I'm feeling empowered!

Sat, 24 Aug 2002 12:21:46 GMT

Research Blogs:

Thu, 22 Aug 2002 01:14:32 GMT

The forces of the web are at work... Online or Invisible reports that papers published on the web are more often cited than papers not web-ified.

Uzilla Scip's Along

Wed, 21 Aug 2002 00:16:16 GMT

I'll be presenting "Uzilla" at the Society for Computers in Psychology in late November. It's in Kansas City just prior to the heavy duty experimental psychology conference, Psychonomics. The next step in this process is a paper for Behavior Research Methods that would provide the critical defining citation for users of Uzilla.