Lynne Kiesling
Over at Crooked Timber, Eszter Hargittai has a post on using social network modeling to make movie, music and book recommendations. I am particularly intrigued by social network modeling, especially the work of Duncan Watts. In addition to analyses of tastes, relationships, organizational structure, etc., I think it may be useful in thinking about institutional change in electric power regulation. But my ideas are still a bit embryonic.
I will try the music links she recommends, and I expect that they will give me better results than the lists that accompany artist listings at allmusic.
Any other recommendations of applications of social network theory to tastes?
This seems mostly unnecessary to mention, but I’ll do it by way of anecdote. During my IT consulting days the company I worked for got really interested in some of this stuff as a way to model the “community” that was being built by various websites. Turns out, a good deal of work on the issue (at least from a modelling standpoint) is being done in computer science departments (which is what I figure y’all already know, but what the hey…).
I’ve not kept up on it much, other than to follow web trends. Not sure if these are of any use or not, then, but here are two I’ve read recently and found interesting:
The Open Source Software Development Phenomenon: An Analysis Based on Social Network Theory
Short and punchy; developers choosing their projects as shown by a power-law model (non-random/preferential attachement).
and
When Being Weak is Brave: Privacy Issues in Recommender Systems.
While not directly on tastes, it does touch on modelling according to jointly recommended items (the width for the hammock jumps) and delves more into weak and strong ties than I’d seen elsewhere (though, that’s not saying much admittedly).