I’m finally starting to make my way thru Programming Collective Intelligence by Toby Segaran. It’s been mocking me from my desk for the past weeks and I wouldn’t let myself pick it up until after a launch and after the holidays. Since both of those have passed (and yes I realize I haven’t spoken about the launch but I have to wait until next week), I can now start reading.

All of his code is in Python which I think is awesome but I’m working in Ruby now. As a way of internalizing the ideas, I’m rewriting everything into Ruby. I’ve pasted the first couple of scoring algorithms. I can’t guarantee that my Ruby is the most elegant way it could be done but it’s working and I’m learning.

The first two ways of gauging the similarity of two movie critics are the Euclidean Distance method and the Pearson Correlation Method. Both seemingly have their pros and cons but the nice thing about the Pearson method is that it can somewhat level out the ratings so even if one critic is constantly harsher than another, it can be balanced out.

• BROWSE / IN TIMELINE
• « Snapshots of the Candidates
• » Newspapers as a Social Object?

• BROWSE / IN Code
• » Starling