Adaptive web sites: A knowledge extraction from web data approach

This book can be presented in two different ways. Based on the title, we can think that this books introduces a particular methodology to build adaptive Web sites. To achieve that, the book first introduces all the concepts needed to understand and apply the proposed methodology.

The alternate view is as a book that presents the main concepts behind Web mining and then applies them to adaptive Web sites. In this case, adaptive Web sites is the case study to exemplify the tools introduced in the text. In the following paragraphs I use the later view to give the reader a quick tour about the book's content.

The authors start by introducing the Web and motivating the need for adaptive Web sites, which is mainly commercial. The second chapter introduces the main concepts behind a Web site: its operation, its associated data and structure, user sessions, etc. Chapter three explains the Web mining process and the tools to analyze Web data, mainly focused in machine learning. The fourth chapter looks at how to store and manage data, as in a popular Web site we can generate gigabytes of logs in a short period of time. Chapter five looks at the three main and different mining tasks: content, links and usage. The following chapter covers Web personalization, a crucial topic if we want to adapt our site to specific groups of people. Chapter seven shows how to use information extraction techniques to find user behavior patterns. The subsequent chapter explains how to acquire and maintain knowledge extracted from the previous phase. Finally, chapter nine contains the case study where all the previous concepts are applied to present a framework to build adaptive Web sites.

In other words, the authors have taken care of writing a self-contained book for people who want to learn and apply personalization and adaptation in Web sites. This is commendable considering the large and increasing bibliography in these and related topics. The writing is easy to follow and although the coverage is not exhaustive, the main concepts and topics are all covered. The level is fair and appropriate for a last year undergraduate or a graduate course.

In a personal digression, I am really glad to see a book where one of the authors was one of my students and, hence, feel that I have contributed with a small grain of sand to this book. In fact, having wrote a few books, I know how hard it is to write one and how much motivation you need to finally see the printed outcome. Finally, without doubt, nowadays e-commerce is one of the main sales channels and clearly learning from your Web customers has a significant commercial value. Sadly, not too many Web sites use the full potential of Web mining and I hope this book makes an important contribution to change that.