This panel defined folksonomy and the current state of tagging. During the course of the conversation there were a few suggestions on ways to improve upon the current state of affairs, which is what the panel was really about. I wanted to expand on that part of the conversation.
It seems to me that every time a cool new technology or idea comes around, we jump on the bandwagon with such enthusiasm that we leave behind everything that came before. When the discussion repeatedly turned to letting non-power-users rely on or draw from the knowledge of the group, either in the form of automatic tagging or keyword suggestions, people questioned how any one person could rely on the opinions of another. While I agree that no system is perfect (systems are, after all, created by imperfect humans), it seems to me that we’ve been developing methods of quantifying trust for a long time.
When I buy something on eBay I’m reasonably assured that I’m buying from a reputable seller. The same goes for opinions on Epinions. Does anybody remember when Epinions was the cool new kid that would empower the masses? I still use the site when making purchasing decisions. For that matter, look at any peer-to-peer network.
Simple ideas tend to be the best ideas. That’s my corollary to Occum’s Razor. Simple methods of user ranking can create a trust mechanism that would allow novice users to rely on the wisdom of experts. This could easily allow for experts in specific areas (Jazz was mentioned during the panel).
I want to repeat one suggestion that was brought up during the panel: tagging should be incorporated directly into the browser (and operating systems). Regular bookmarking should take advantage of tags. This should not replace categorization but rather be offered as an option to it. Just because you think it is better doesn’t mean it has to replace something people have already learned to use. There is something to be said for the efficiency of knowledge already learned.
That suggestion led directly to something that I have not heard mentioned by anyone despite it being extremely obvious. In fact, it hadn’t occurred to me until the panel. Tagging has been around since the early days of the web. It has been done by the experts in each subject area and it has been done to a vast quantity of what is out there. I’m talking about keywords. Remember meta-tags?
Why has nobody created a bookmark plugin or web utility bookmarklet that automatically includes meta-data with the link URL? This should be automated in every browser and bookmarking website and, at the very least, include the two most common meta-tags: description and keywords. Relying on past lessons learned: bookmark searching should be able to very easily include or exclude metadata in order to deal with keyword spamming (there we go re-using past knowledge to enhance the trust of the system…).
I’ll leave off there. Hopefully other attendees (and the panelists themselves) can offer more suggestions or point out the flaws in mine. The evolution of ideas is best accomplished through collaboration.
Related Links:
Panel Listing on SXSW
Beyond Folksonomies (great resource page)
PREVIOUS: Countdown: One Day to SXSW | NEXT: SXSW Update
I was at the panel discussion. The talk seemed unfocused and was a bit of a let-down to be honest. Starting from the beginning with the struggle over the very definition of “folksonomy” to the analogy of tags and folksonomies being “pigeon english”, I thought there would be more focus on where things are going and usability.Would have been nice to see more discussion on approaches to collaborative annotation that exist in between macro – and – micro views of data.
Jon, I agree completely. Because I was cross-posting to the community blog I decided to not to be critical of the panel discussion and to just try to move the conversation forward.