Dernière mise à jour : 11/03/2009 - 1 livres - 1 critiques
  |
Public visé : avancé Résumé de l'éditeur
Product Description
There's a great deal of wisdom in a crowd, but how do you listen
to a thousand people talking at once? Identifying the wants, needs,
and knowledge of internet users can be like listening to a mob.
In the Web 2.0 era, leveraging the collective power of user
contributions, interactions, and feedback is the key to market
dominance. A new category of powerful programming techniques lets
you discover the patterns, inter-relationships, and individual
profiles-the collective intelligence--locked in the data people
leave behind as they surf websites, post blogs, and interact
with other users.
Collective Intelligence in Action is a hands-on guidebook for
implementing collective intelligence concepts using Java.
It is the first Java-based book to emphasize the underlying
algorithms and technical implementation of vital data gathering
and mining techniques like analyzing trends, discovering relationships,
and making predictions. It provides a pragmatic
approach to personalization by combining content-based
analysis with collaborative approaches.
This book is for Java developers implementing Collective Intelligence
in real, high-use applications. Following a running example
in which you harvest and use information from blogs,
you learn to develop software that you can embed in your own applications.
The code examples are immediately reusable and give the Java
developer a working collective intelligence toolkit.
Along the way, you work with, a number of APIs and open-source
toolkits including text analysis and search using Lucene,
web-crawling using Nutch, and applying machine learning
algorithms using WEKA and the Java Data Mining (JDM) standard.
About the Author
Satnam Alag, PhD, is currently the Vice President of Engineering
at NextBio, a vertical search engine and a Web 2.0 collaboration
application for the life sciences community.
He is a seasoned software professional with over fifteen years
of experience in machine learning and over a decade of experience
in commercial software development and management.
Dr. Alag worked as a consultant with Johnson & Johnsons's BabyCenter
where he helped develop their personalization engine.
Prior to that, he was the Chief Software Architect at Rearden
Commerce and began his career at GE R&D.
He is a Sun Certified Enterprise Architect (SCEA) for the Java Platform.
Dr. Alag earned his PhD in engineering from UC Berkeley
and his dissertation was in the area of probabilistic
reasoning and machine learning. He has published
numerous peer-reviewed articles.
Critique du livre par Chabli Faisel
A mon avis, ce livre est destiné à des gens ayant de bonnes connaissances en Business Intelligence ainsi qu'en Java pour bien comprendre les exemples fournis, mais qui reste un livre pratique pour l'application de l'intelligence collective à des applications web.
Etant divisé en trois grandes parties, le livre nous introduit, à travers le 1er, 2ème et 3ème chapitres, au domaine de l'intelligence collective en nous donnant un aperçu de l'architecture nécessaire pour intégrer l'intelligence collective dans une application.
La 2ème partie du livre traite l'intelligence provenant de la collecte de données.
A travers le 7ème chapitre, le lecteur est introduit au processus du data mining ainsi que ses différents
types d'algorithmes. L'outil utilisé étant WEKA, un outil open source pour
le data mining dont l'utilisation sera plus explicite dans le chapitre 9 à travers
des exemples de clustering.
La dernière partie traite l'application de l'intelligence dans une application.
Les points forts du livre résident dans la méthodologie adoptée par l'auteur pour expliquer comment appliquer
de l'intelligence collective dans une application web. Ainsi, le livre
nous mène, à travers de nombreux exemples, à comprendre comment appliquer
de l'intelligence collective, en dénombrant ses bénéfices,
dans une application web. Chaque concept est ainsi illustré par
des exemples concrets. L'auteur cite plusieurs
sites de référence utilisant le concept de l'intelligence collective, chose qui peut donner
plus de visibilité aux lecteurs désirant appliquer ce concept. Une dernière chose que j'ai beaucoup apprecié dans
ce livre est qu'il nous transporte du monde du technique abstrait
vers un monde du pratique en prédisant le besoin des internautes.
Certains aspects techniques présentés par le livre peuvent causer la perte
du fil conducteur pendant sa lecture tels que la présence
de plusieurs diagrammes de classes et du code Java. J'aurais
préféré que le livre ne traite pas les aspects conceptuels
des choses à travers du technique.
Ce même problème est rencontré par la présence de certaines formules
mathématiques, J'aurais préféré que l'auteur propose des moyens faciles
à mettre en œuvre pour aboutir à des résultats exploitables
si c'est nécessaire de passer par des chiffres (calcul de la corrélation, similarité,…).
Ce que je n'ai pas aimé dans la méthodologie de l'application du data mining
dans ce livre est le fait d'utiliser directement l'API Java de WEKA
au lieu d'utiliser l'outil complet WEKA. Ceci, afin d'éviter
de réinventer la roue. Aussi, l'auteur devrait normalement
expliquer comment sélectionner les attributs de prédiction
pour la construction d'un modèle prédictif.
English version :
In my opinion, this book is intended for people with good knowledge
in Business Intelligence as well as Java to understand the examples
provided, but this is a practical book for the application of collective
intelligence in a web applications.
Being divided into three main parts, the book introduces us,
through the 1st, 2nd and 3rd chapters, to the field
of collective intelligence by giving us an overview
of the architecture necessary to integrate collective intelligence
in an application.
The 2nd part of the book deals with the intelligence derived
from the collection of data. Through the 7th chapter,
the reader is introduced to the process of data mining and its
different types of algorithms. WEKA tool was used, an open source
tool for data mining whose use will be more explicit
in Chapter 9 through examples of clustering.
The last part deals with the application of intelligence in an application.
The strengths of the book lie in the methodology adopted
by the author to explain how to apply collective intelligence
in a web application. Thus, the book explains, through many examples,
how to apply collective intelligence, by enumerating its profits
in a web application. Each concept is illustrated by concrete examples.
The author cites several notable sites using the concept
of collective intelligence, something that may give more visibility
to readers who wish to implement this concept. One last thing
I greatly appreciated in this book is that it brings us from the
world of abstract art to a world of practice in predicting the need
for Internet users.
Some technical aspects presented by the book may cause
the loss of the thread during playback, such as the presence
of several class diagrams and Java code. I would have preferred
that the book does not address the conceptual aspects
of things through the technique. The same problem
is encountered by the presence of some mathematical formulas,
I would have preferred that the author offers easy ways
to implement in order to achieve useful results if it is necessary
to go by numbers (calculation of the correlation , similarity, ...).
What I did not like in the methodology of applying data mining
in this book is the use of an API directly from Java WEKA
instead of using WEKA tool . This is to avoid reinventing the wheel.
Also, the author is expected to explain how to select
attributes predictive for the construction of a predictive model.
Sommaire
425 pages,
28/09/2008
Editions Manning Publications,
ISBN10 : 1933988312 |
Copyright © 2009 Chabli Faisel. Aucune reproduction, même partielle, ne peut être faite de ce site ni de l'ensemble de son contenu : textes, documents, images, etc. sans l'autorisation expresse de l'auteur. Sinon vous encourez selon la loi jusqu'à trois ans de prison et jusqu'à 300 000 € de dommages et intérêts.
