Building Bubble-Up Folksonomies

by Joshua Porter  |   September 12th, 2005  |  shortlink: http://bokardo.com/p/209

Tom Coates writes up how to build bubble-up folksonomies. It’s an interesting piece, sure to become more useful over time, as systems migrate toward bottoms-up instead of top-down.

Are you thinking of building something like this?

Make them Care! - Struggling to communicate the value of your product or service? I'm writing a new book that shows you how to make people care about your product or service by clearly communicating the most important bits. For designers and marketers creating product web sites. Find out more.

Links to this Post

Comments

1.  CM Harrington 2:25am, Tue 13th, 2005

To be honest, I am not entirely sure I understood the concept. It sounded like this: Everyone sees a thing. Everyone tags it with whatever they want. You take the most popular (x) tags, and kill the rest. *BAM* there’s your “fixed” tag association (“fixed” being in quotes because I would imagine it’s meant to change over time).

But that doesn’t change the fundamental problem with public tagging. Actually, upon further inspection, it exacerbates the problem. The problem being not everyone conforms to “the mob”. For example, I have this nasty tendency to spell things in UK English, despite the fact that I’ve been in the US for 17 years. If I try to look for an article by tag, chances are pretty good that I won’t have the foggiest notion how to go about doing so, purely on the basis of spelling. Not only that, but I don’t think similarly to the vast majority of people on the planet, so chances are pretty good that a tag would be obvious to me, but unthought of by others (and vice versa).

Tags/folksonomies/etc. are only good in two areas: Personal use, where you build your own tag system; and when you want to surf for information/serendipity (“oh, what does this link give me?”). However, tagging systems are *really* quite bad when you are searching for specific data, and you want the fastest route possible in a given interface.

2.  Tom Coates 7:00am, Sat 24th, 2005

No no – that’s not it at all. The point is that some things are made of collections of smaller things, and that if you get people to describe the smaller things then you (often) end up with a set of data that you can derive descriptions of the larger things from too. It won’t always work – some collections of things work dramatically differently than their component parts – but for some things (like describing musical styles) you can go – in the simplest example – ‘these songs are all pop, so the album they all appear on is probabaly a pop album, and the artist that played them is probably a pop star’.

You’re completely right that tags are about discovery rather than goal-oriented actions where only one piece of data will do. But that’s a useful goal and the distinction isn’t really relevant here. This is a way of finding a way to help people explore a musical space by making it possible for them to move between songs, artists, albums, radio shows and back to other songs, by making best use of the data you collect in one particular territory.