Tagging is all the rage. It’s the marquee feature of WordPress 2.3, and it ties into all kinds of cool trends in information organization and visualization. Just to make sure everybody’s on the same page, Wikipedia defines the particular kind of tagging of which I speak thusly:
A tag is a (relevant) keyword or term associated with or assigned to a piece of information (e.g. a picture, a geographic map, a blog entry, or video clip) as a whole or only to a part of it (e.g. “timed tags” assigned to specific moment in time in a video), for purposes of keyword-based classification and search of information.
Being a visual guy maybe a bit more than a word guy, I didn’t fully grok the greatness of tagging until I discovered the tag cloud. A tag cloud is what you get when you take a bunch of tags for a particular group of tagged items and display them together while showing the frequency (or some other attribute I guess) of each tag with text size, color, position or some other visual attribute. Soon this blog will have a tag cloud.
In fact, tag clouds have displaced Edward Tufte’s sparklines as my favorite information visualization tool. But I still have to recommend Bizzantz’s sparkline tools. I wish I could have used them more at work.
I had to make one of my own. A quick search led me to TagCrowd, a tool that turns any list of words into a tag cloud. But I didn’t have any blobs of tagged data, so I took the first thing at hand, the text of a web site I was viewing, and dumped it in there. Here’s what I got:
admin administration agreement anti-virus base blog buy client customer demo email feedthe file ftp ftpplanet gold imail instant ipswitch learn license managed messaging monitor network news offer overview partner policy products professional program reseller save secure server service software special ssh ssl support trademarks transfer v11 video2-minute view whatsup ws
created at TagCrowd.com
Pretty interesting. Again, it’s not really a tag cloud, but it is a word cloud of some sort, and it does give you a pretty good idea of what the source material is about. You can probably even guess what site it was.
There are definitely some flaws in TagCrowd’s engine and even in the whole idea of using type size to indicate information when the words are of different lengths and densities themselves. For example, which of these is in a bigger font?
In any case (sic), I enjoyed this reductionist method and decided to apply it in a thoroughly inappropriate setting. I made a tag cloud from the text of Walt Whitman’s When Lilacs Last in the Door-yard Bloom’d
amid bird black blossoms bush carol cedars chant cities coffin comrades dead death debris door-yard droop fields gray-brown green hands hear holds joy leaves life light lilac love night passing pines recesses rest rising shall sing sky song soul spring star strong suffer swamp thee thou voice walk west western
created at TagCrowd.com
I’ll admit two things here. One, I know this is an awful thing to do to poetry. And two, I stacked the deck by using Whitman. This cloud is a decent approximation of the themes of the poem, largely because of the repetitive nature of Whitman’s incantations and the length of the poem. I’m not planning to try this with Bassho or Emily Dickinson.
I really like the look of this as typographic design, and I hope that TagCrowd enhances their tool or that I find another tool or method with more type options. I’m still thankful that they make it so easy to copy the HTML for these clouds.
Returning from that poetic reverie, here’s a little more from Wikipedia on tags:
Tags can be assigned to a piece of information by its author/creator or by consumer/viewers/community. Tags are typically used for resources such as computer files, web pages, digital images, and internet bookmarks (both in social bookmarking services, and in the current generation of web browsers – see Flock). For this reason, “tagging” has become associated with the Web 2.0 buzz. Many people associate “tagging” with the idea of the semantic web, however some believe that tagging may not be having a positive effect on the overall drive towards the semantic web.
That’s where this stuff gets even more broadly interesting. The idea is that if enough people tag enough stuff, we’ll have a layer of understanding about content far beyond what spiders and bots can gather. Some say this will only work if people are good at tagging, but I’m thinking this is a wisdom of crowds kind of thing – even if individual taggers don’t know much, the mass of them know a lot. And that’s what a tag cloud can help show. I hope.