I am a member of Poznan Java User Group. We have active book review program with O’Reilly, APress and Manning. I wrote some reviews and I’ll post them all, today my first one, “Collective Intelligence in Action”.
Collective intelligence is very popular these days. Thanks to the Internet companies we can use this concept every day. I have to admit that I am a big fan of “intelligent” services. I search news on Digg, I listen to music through the Last.fm player, I use Wikipedia, YouTube, Amazon etc. everyday. I have been wondering many times how these sites work. The curiosity led me to read “Collective Intelligence in Action”. After I read it, I was surprised that this book is so practical. Theory is limited to minimum. After reading it, you should be able to add the CI features to the existing sites.
Author, Satnam Alag, has organized his work in a perfect way. Every chapter has an introduction, a summary and very handy references – I used them many times. All the mathematical concepts and definitions are shown in examples. There are lot of a Java code listings, therefore the basic knowledge of this language might be useful. The theoretical foundations are not necessary. Each chapter can be treated separately, but together they create coherent paper about recommendation system. The author of the code takes care not only about the correctness but also about the efficiency and the scalability.
The fact that Satnam presents a lot of stable and useful open source software is worth of noticing. Projects like Nutch, Lucene or Weka can be easily adapt to our services. The book shows how to do it from programmer’s point of view (API).
I recommend “Collective Intelligence in Action” to the Java developers who would like to know how to build recommendation systems, intelligent search of theirs resources, automatic tagging or network crawling. Book is worth reading even if you do not plan to use CI in your application. Base Web2.0 mechanisms are very easy to implement and do not require a lot of theoretical knowledge, Satnam Alag has proven this in his paper.
I would not recommend this book to data mining or text analysis experts. This is not an academic work, people who are looking for theoretical information about CI could be disappointed.
Tags: book review, collective intelligence, java