More Information on my Book
I mentioned a couple of months ago that I would be writing a book for O’Reilly (the publishing company, not the TV host). Now that I’ve achieved my first writing milestone, and have a few chapters completed, I can share a little bit more.
It doesn’t yet have a confirmed title, but the subject of the book is “Collective Intelligence”. This is a phrase that’s getting a bit more use lately and probably encompasses a number of meanings, but in this context it refers to taking the behavior or preferences of a group of people and combining it to create new information. A familiar example, and the subject of the first chapter, is a recommendataion system like Amazon’s, which uses the purchases of thousands of customers to guess what another customer might like to buy next. Other topics include automatic document filtering, search engines, online dating, blog groupings and interpreting server logs.
The book explains a number of machine learning algorithms and statistical methods that are used to combine the information. In contrast to many AI books that I’ve read I’m trying to give a lot of real-world examples related to internet applications and provide complete working code so that readers aren’t stuck trying to turn abstract ideas into a program. Many of the examples involve combining a machine-learning algorithm with API calls from another website such as Google or Kayak. The code is in the book is in Python, and I’m hoping to provide the examples in several other programming languages online after the book is published.
Anyway, I hope this subject interests you or someone you know. I’ve been noticing an increasing interest in these types of techniques in a variety of fields — I’ll talk more about that in a later blog post.