Thomas Vander Wal wrote as a comment to How Buildings Learn, for social software
This idea of ease of finding people to talk to around popular books, but difficulty finding more niche books is something the dating tag site Consumating.com called quirkiness. Ben Brown explained they had a measure for quirkiness that surfaced quirky connections between people. It was just above outliers to a few 10s higher. With 300k people in Consumating the quirkiness factor ran from 7 to about 40.
Quirkiness was people who had relatively rare tags in common. This rare commonality was something that was really difficult to find in the wild. This is one of the benefits of using digital means to connect people. Consumating found the relationships that were lasting quite often were grounded in this quirkiness. This came up on a panel I was on w/ Ben Brown, which was moderated by Heath Row who met his wife on Consumating as they were both Manhattanites who were tagged “mountain climbers”, hence quirky.
This is the drawback of popularity-based recommendation systems. Sometimes they tell us things that are new and hot that we haven’t seen already. But often they tell us things we already knew. If you liked Harry Potter Book 5, you might like Book 6. Statistical improbability, social filters, and the combination of the two can lead to more interesting results than popularity alone.
Comments on this blog are broken, and will stay broken til I upgrade to the next version of MT or to WordPress.