Weblog
Oct 28, 04:35 am: Loosely coupled communities and the nptech tag
At this weekend’s Tag Camp, I’m going to be talking about the surprise (at least to me) of the nptech tagging experiment. And that’s the way that helped provide a community badge or identity.
Let me flesh these thoughts out a bit:
- In December, 2004 a conversation about Nonprofit Technology Taxonomy Creation began on the Omidyar Network
- Let me interupt to say that trying to get a discussion going on “Nonprofit Technology Taxonomy Creation” made me want to empty a staple gun into my forehead
- Luckily, that’s not the direction the conversation took. Instead, we decided to use del.icio.us and a (relatively) opaque tag and see if we could parse a taxonomy out of the results.
- We started spreading the word a little later that month
- The idea, at this early stage, was that we would collect enough links and then analyze the tags and find out the best words to use in a taxonomy—floss, foss, f/loss, oss, open source—the tags would tell us which to pick as the main word and then how to stem the other words someone might use in a search
- That was the idea
- That was very hard
- For a couple of reasons: the nptech tag itself was opaque enough that people kept adding “nonprofit” and “technology” and leaving off any more granular tags
- Damn them using it for their own recall
- And the API didn’t let us get at the data (nor did we immediately) have enough to help us get deeper into any kind of emergent taxonomy
- We first got data (courtesy of Brian Del Vecchio) in February, 2005 and found that 31 users had tagged 470 unique links
- By June 2005, 124 users had bookmarked 1214 unique links.
- As of today (again, courtesy of Brian Del Vecchio) 226 users have bookmarked 1976 unique links
- And they’d bookmarked blog posts in technorati
- And events in upcoming
- And more links in furl
- And created a playlist
- So what that seems it indicate is that people latched onto the tag as a badge, a way of creating community
- Does that seem right?
- Peter Campbell put up a site to aggregate the various feeds
- And try to make the community more tightly coupled
- But we haven’t given the site the attention to get any momentum going there
- So what’s next?
- I’d love to get data out of all the tag systems about the users
- Even more I’d love to see if there’s a way to determine the relationships between the taggers
- That’s more interesting now than the tags themselves
- How can we do that?
- And a survey. It’s been sitting, unpublished, in the community site
- What should we try to find out about the users to try and understand what if any community exists?
- What conclusions?
- Tagging can support information sharing in loosely coupled, distrbuted groups that operate in a leaderless fashion to a much greater degree than can things that require a higher cost (list servs or community message boards)
- How can that learning be used? Particularly in a nonprofit context.
