Visualizing wikis: History Flow

It’s all in the visualization.

When I describe wikis to someone still grappling with the oddness of the word ‘wiki’, my description inevitably kicks into abstractions about joint authoring, organic development, networked interactivity. What is likely to lodge, in an innocent auditor’s mind, is an amorphous sense of wide-open vulnerability: You mean, anyone can change any page? – or Why would anyone want to risk having their work distorted/mangled/deleted?

Assurances about communal good will and self-policing don’t always reassure. It’s a world of vandalism, terrorism, and error; the Internet is never secure, the Web is never authoritative; without accountability there is no truth… und so weiter. Descriptions of ‘rollback’ functionality are more comforting to our skeptical strawman, but of course that’s just damage control.

So here’s a better picture. Back in 2003, the Collaborative User Experience Research Group at IBM began publishing some illustrations of wiki posting activity. Authors were assigned a range of colors, in order to track the influence of any given individual, revision activity, and information persistence in a collaborative environment. This tool, called History Flow, is now available as a download at IBM’s alphaWorks.

These are visualizations that are immediately graspable. Here’s an example of three persons contributing to one wiki page over the course of four versions. Their words are color-coded:

And here’s the same content, now shaded to show persistence. The oldest surviving content is darkest:

This all gets very interesting when the researchers visualize the activity of a particularly controversial page, such as Wikipedia’s entry for abortion. This page does indeed attract outright attacks and vandalism; it’s a wonder that such a contentious topic is at all viable as an ‘open’ forum. But History Flow shows how quickly a wiki can recover from sabotage – “so quickly that most users will never see is effects” (text and illustrations here). Here’s a chart of the ‘abortion’ page, charted by saved versions; the gaps represent deletions:

And here’s the same page, now represented across a time axis: the attacks on this page are, by this measure, pretty much undetectable:

May History Flow keep flowing, keep rendering snapshots of how wikis actually work. They can be reassuring, arresting, even beautiful, given the right colors.

A 2004 report on History Flow entitled Studying Cooperation and Conflict between Authors with History Flow Visualizations, by Fernanda B. Viegas, Martin Wattenberg, and Kushal Dave, is posted here.

ALA | Intellectual Freedom

ALA | Intellectual Freedom

Row row Rosie boat

Scott & the ladies go cruising – along with 500 other gay families. All aboard!

Library Success: A Best Practices Wiki

Library Success: A Best Practices Wiki

Selection sources for college library e-resources

1. ACRL Internet reviews, listed chronologically, and the ACRL Index to Internet Resources, grouped by topic.

2. Choice reviews (GSLIS access only), advanced search: electronic format, outstanding titles, last five years.

3. ARBA online (GSLIS access only), ‘website’ subject search

4. Keeping up with the Joneses: Dartmouth Library’s Recent Digital and Electronic Acquisitions

5. Electronic Books and Texts: Case Studies and History, a section of the Scholarly Electronic Publishing Biobliography

6. A few journals: Journal of Digital & Electronic Acquisitions, Journal of Digital Information, Journal of Digital Contents

The Chronicle, 7/15/2005: Romantic Poetry Meets 21st-Century Technology

This week’s Chronicle of Higher Education writes up RAP. A relatively thorough treatment – kicks off with “dumb luck” and goes from there.

There’s a little discussion board accompanying the article. My favorite post so far: ‘English prof’ writes: “As for, ‘And some enthusiasts say that the technology can actually change students’ writing for the better, by encouraging them to swap ideas with their classmates and to revise their work continually’ — we can do that sitting around a table together, talking, and having lemonade.”

Ahh, sweet cold lemonade. The impeccable drink…

Her lover brings the lemonade, she sips;
She then surveys, condemns, but pities still
Her dearest friends for being dress’d so ill.
Beppo LXV

North Haven 2005

In which someone turns 40, someone turns 39, someone works the road & refuses to wave. In which Maine is warmer & we sing less but dance more than in years past. In which the sky is clear and bright all weekend, & we once again escape the Fourth.

TagCloud

Folksonomy meets RSS: TagCloud.

Yes I know: ‘folksonomy’ is a particularly hideous neologism – it basically means metadata assigned by a non-hierarchical community. D-Lib let the word rattle and clunk around a rundown of ‘social bookmarking tools’ in April. Ever since it’s been nagging me – it’s just too unholy of a mix of populism and taxonomy.

TagCloud describes itself as an “automated folksonomy tool” – which I think is a particularly trendy/tortured way of saying: it scans RSS feeds for frequency of terms, then display the most prominent terms in text that is scaled to represent frequency. You know, like Flickr’s tags.

It’s a work in progress, and Danny Sullivan has written about some of its limitations in SearchEngineWatch. But TagCloud seems like a good way to get a torrent of RSS-fed information into some kind of visual handle on the user end. You know, where folks live.

Wikinews

Tail-chasing in news: after running across a story in the Boston Globe about the inevitable newWikinews (actually, after enduring a splash ad & then running into the story), I wondered how a venture like this (so amateur, so unauthoritative, so doe-eyed Internet) would actually work.

So I picked a story that might well attract obliteration of just-the-facts-ma’m: Wikinews coverage of the Supreme Court uanimous slapdown of file sharing – with the intention of comparing it to ‘real’ news coverage.

Surprisingly, the New York Times had nothing about this story on its home page – the NYT was consumed, instead, with the covering the Court’s refusal to consider saving Judith Miller from going to jail for protecting her sources.

So I next turned to CNET’s coverage of the file-swapping ruling . It was a more extensive than the Wikinews treatment – but, as of right now, the tone seems similar; both writeups of the story are balanced and neutral in tone. Of course, the Wikinews entry could and will change.

Fluidity make a return visit to the same story on Wikinews interesting – this kind of news is its own development. And, at least in this case, it swamps the ‘story’, newssharing over filesharing.

Urban Design: A Research Guide

Urban Design: A Research Guide