Alla Zollers notes that we can tell a lot about a community by looking at its tags. E.g., delicious.com is clearly IT heavy. “Relevance is socially constructed,” she says. I.e., what is relevant changes over time depending on who interacts with it and how they interact it. “Communities form around tags.” [As always, I’m just taking live notes and missing lots.]

Lilly Nguyen thinks of tagging as peer production of knowledge. People think about taggging in terms of production but not use. We have sociological tools that explain some of these but they don’t seek up with Library Information Science. Use of tags is not just about findability and refindability.

I say that I agree that finding and refinding doesn’t cover it. I sometimes tag because I feel like I’m contributing to a tag stream that’s creating knowledge, along with all the human motivations we have for doing things (reputation building, validation, etc.).

Barbara Wildemuth asks why people would use (not create) tags other than finding and refinding. Lilly says she uses tags to discover, to browse, which is different than finding.

Lilly says she worries about tagging overloading us. Someone says we’ll fix it eventually. Lilly says, “So fix it!”

Thomas says that he feeds delicious.com for social reasons—he knows people pay attention to it—but he searches through Yahoo MyWeb because it has features that work better.

Cliff Lampe wonders where the line is between finding systems and recommendation systems. (Thomas likes Ma.gnolia.com.)

Gary Marchionini points out that people tag also because the act of tagging helps memory, just as taking notes does.

Thomas points out that people tag with terms they’ve most recently interacted with.

Jackson Fox says that we transition terms over time. E.g., after tagging pages as “web design” for a while, he started instead tagging some “graphic design.” So, when he goes to find info about color theory in ten years, if he doesn’t remember he used to call them all “web design,” he’ll miss them. The tools don’t handle this well, he says.

Now Jackson gives his official talk. He says content is very personal. When content creators tagged their own material at lulu.com (where he works) and found out that others could see those tags, they were livid. Then they calmed down and saw they could connect their content with others. Is it a folksonomy if the taggers are the content creators? How do you avoid spamming?

Thomas: Creator’s tagging doesn’t fit within folksonomy, but it can be a type of seeding.

I say that if a site made the author’s tags visible next to the readers’ tags, it’d stimulate anti-author tags.

Jung Sun Oh says in her talk that we don’t know whether people tag based on their own mental models or based on the terms in the source they’re tagging.

Terrell Russell says he’d like to help people figure out what they know by seeing one another’s tags. If you make the tags and concepts visible and put them in the middle the room at all times, you’ll begin to converge. Thomas says Cameron Marlowe at Yahoo has built something like this. He says that in one instance, people were removing tags about themselves. Nicole Ellison says that in some organizations, you don’t want to be tagged a particular way because it might drive people toward you. Zeynep Tufecki says that people care a lot about their identity; she’s reminded of research about identity in prisons and other places where you don’t have control over that. Jackson says that on the other hand, you can build social capital that way. Amanda Lenhart points to the identity brouhaha at FaceBook as an example of the dangers. Cliffe Lampe suggests pairing it with a reputation system. There’s discussion of whether such a system would result in only positive assessments. Lilly worries about mistakes staying with your forever; Terrell says he’s working on tag decay. Fred Stutzman says this is a social attribution system, but tagging boils things down too much so it’ll be harder to discover what we don’t know about a person. Terrell says that he’s aiming at skills areas. Laura Sheble says it might be helpful to have a way to respond. Also, it’ll be interesting to look at the edge cases, she says. [Tags: unc taxonomy folksonomy everything_is_miscellaneous ]