The author present us with a gentle and light introduction to collective intelligence. Also, he showcases many examples readers are most likely familiar with.
I think some of the most exiting examples are Netflix and eHarmony. These are two companies using lots of information to predict and infer new information to attain two very different goals. One pure entertainment and the other possibly very serious.
The book has no exercises with datasets from either of these services. Netflix, however, does provide some ‘mine-able’ data as they sponsor a contest in the hopes of continuously improving their recommendation algorithms. Another good example of this is Amazon’s recommendation system.
I’ve read about the Netflix contest in the past but I am not too familiar with the current standings. Chapter two goes into a fairly complete example of movie recommendation systems such as Netflix. It would be really fun if some dataset like this (of movies) were available for books. Imagine having book recommendations for those you care. Better yet, knowing what to recommend to someone with a specific need.
Author does skim over the limits of some of the methods employed in intelligent systems as well. One of the side effects of these systems that I am familiar with is that, every time I go to Amazon, I get recommendations (good ones) for items which may be of interest to persons I’ve bought something for in the past. In this case, the data gathered and predictions are good but, the problem domain is very big since Amazon would have to keep tabs on everyone you ever buy something for. Now that would be some amazing data to use…
Al in all, I can’t wait to start going over the examples, maybe even the exercisers in the back of every chapter. The author decided to use Python and I think its great. I have no Python experience and am really exited to give it a shot. Code samples look fun and primer on the web show Python to be a very practical choice for this book.