Recently I have been thinking about how we can apply the techniques of collaborative filtering to people search and recommendation. Indeed we have built Beanbag Learning as a research testbed to allow us to explore possible approaches in this space. There are a number of threads which are starting to come together: collaborative filtering, the social graph, peer to peer, Whuffie and solar magnitude, openness and mutability. I’ll outline them below and try to link to some of the wide range of conversations going on which link these seemingly disparate areas.

Read the rest over on the Scrump blog »