Trust and Folksonomies

In a job interview two weeks ago, I was asked a question about the potential for malicious tagging if libraries opened up their catalogs and allowed users to tag books as they pleased. At the time, I answered by saying that most tagging systems handled malicious tags by ignoring them: if only one person tags an item with a malicious tag in a system like LibraryThing that has thousands of taggers, then that tag will sink to the bottom and the more common (and therefore more useful) tags will rise to the top. But after chewing on the question for awhile, I think I want to take that answer back.

Oh, it's a perfectly accurate answer, don't get me wrong, but I want to contest the very premise of the question. Why don't we as librarians trust our users to act responsibly in the library catalog? Consider all of the trust that we already put in our patrons. We let them wander around largely unsupervised in buildings containing millions of dollars worth of books that people have spent hundreds of hours painstakingly shelving in the proper order. We let them check out hundreds of dollars worth of books and DVDs at a time and take those books out into the world, where we have no control whatsoever over what they do with those items. And, really, how often do our users do anything truly catastrophic when given this trust? Yes, there are accidents sometimes, and coffee gets spilled on books or dogs them chew up, but I'm talking about really large-scale, intentional attemps to be destructive or malicious. Have you ever had a group of users, say, decide to re-shelve a section of books by color? (If I was a bored freshman looking for some sort of mildly amusing prank to pull, that's the kind of thing I'd consider. Think how pretty a rainbow of books would be!) How often do bored students pull the keys off the keyboards on the library computers and put them back on in alphabetical order? (Personally, I don't understand the amusement value of that one, but it was popular with kids in my computer programming classes in high school.) Often enough that you would call it a problem and consider taking some sort of step to cut down on it? If not, then why do you think your users would be any more likely to vandalize the catalog than they are to vandalize your physical facilities?

A New Use for Wikis

I haven't decided yet if this is sheer brilliance or absolutely insane. On the one hand, bringing more transparency to the process of drafting laws, and increasing the ability of non-lobbyists to have an influence on that process, are unquestionably good things. Heck, forget for a minute about the ability of common people to suggest edits to a draft law—can you just imagine all of the voters being able to look at the "edit history" of a law and see what parts were inserted by which people when?

On the other hand, a big part of what makes Wikipedia work is that is has a dedicated core of editors who are interested in truth and balance over any partisan viewpoint, and those editors out-number and out-clout the partisans. This situation is much less likely to hold in a sphere that is partisan by definition, such as drafting laws. As Michael Mussa (a former economist with the International Monetary Fund at a level where even the economists have to be politicians) once commented, "In Washington, truth is just another special interest, and one that is not particularly well financed."

New and Notable Digital Collections

Three new initiatives to digitize information and put it online have been in the news lately.

1) The Boston Library Consortium, which includes most of the major New England colleges and universities (Brandeis, Brown, MIT, Tufts, and quite a few more) is working with the Open Content Alliance to digitize public domain materials in their libraries.

2) The papers of former Supreme Court Justice Harry Blackmun are being digitized and put online. (Hat tip: Volokh Conspiracy.)

3) Robert Heinlein's papers are being digitized and put online (although, unfortunately, you have to pay to get access to them). (Hat tip: Slashdot.)

Link Roundup

A New Approach to the Semantic Web

More about why most published studies are wrong, and problems with the theory.

Some Library Journal love for tagging.

Measuring the Success of a Classification System

A New Toy for Hard-Core LIS Geeks

From Google again, of course.

An open-source semantic topic analyzer for print documents.

Google Strikes Again!

I can only assume that they're planning to launch a U.S. version of this sometime before the campaigning for 2008 really gears up.

Another Player in Semantic Search

A startup called Powerset debuted a new system of semantic natural-language Web searching on September 17. It's currently in a closed alpha, so I haven't been able to poke at it, but it sounds both ambitious and wonderful. Plus, it's build by PARC—the same folks who invented computer mice and GUIs. Here is the page for PARC's natural language processing research project, which is the technology used by Powerset.