PRObooks
Building Intelligent .NET Applications
By now, most of the beginning and even intermediate .NET books that have offered ample training for applying .NET technologies in learning and real-world business scenarios have been released upon the marketplace. (Technical book publishers are rejoicing with the advent of the .NET Framework 2.0 release because they can revitalize their catalogs with updates to their existing inventory of Microsoft technology titles.) Occasionally, however, an advanced book on leveraging .NET technologies answers the call of developers asking for something a little more interesting. Building Intelligent .NET Applications is such a title.
Author Sara Morgan Rea describes herself as the type of developer who is always pushing the envelope looking for more. Developers who seek the same technical experience as Rea will unquestionably be satisfied with the guidance she provides in her exploration of Artificial Intelligence (AI) using the .NET Framework.
The book begins with a very brief introduction to what AI really is all about, from its historical relevance to its functional, albeit unglamorous, role in today s sophisticated applications. Chapter 2 walks readers through the construction of creating a speech application using Microsoft Speech Server and the related SDK. Chapter 3 builds on this knowledge with the construction of a telephony application, and concludes with an example of a multimodal speech application in the following chapter.
The next three chapters demonstrate data mining with an evolving database via the creation of a rule-based application. Chapter 8 instructs readers on how to build agents, from the client-side, much maligned Microsoft Agent characters to server-side Web service agents. Each of these application examples includes real-world case studies, proving the meaningful usage of these concepts in day-to-day business. The final chapter considers the future of AI and the role Microsoft might play in the further evolution of the field, with a majority of advancements expected to emerge from Microsoft Research (several active MSR projects are referenced). Other impending Microsoft technologies are mentioned in their context to AI as well, from Analysis Server 2005 to Longhorn s Avalon, Indigo, and WinFS features. The book concludes with one of the better glossaries I have seen in quite some time.
The book failed to achieve a perfect rating because it lost points on its price per page ratio and it failed to spend more real estate on the other areas of AI that were briefly mentioned in the last chapter of the book. Granted, Artificial Intelligence titles are historically high-priced because of their smaller audience and demanding concepts but given that the book is mostly a collection of broad overviews of AI applied to .NET scenarios, it is $10 overpriced in my opinion. Had the publisher bundled a CD-ROM with the code samples, whitepapers, Web bookmarks, and commercial and open source AI applications mentioned in the book, the additional cost would have been justified. The author/publisher also could have provided additional chapters and coding examples on the other AI areas, including Fuzzy Logic, Game AI, Genetic Programming, Natural Language Processing, Neural Nets, Machine Learning, and even Robotics (programming a Lego Mindstorm brick using a custom .NET assembly would have been an entertaining and educational exercise).
Nevertheless, for those .NET developers ready to expand their horizons beyond the world wide web of radio buttons and checkboxes, I recommend Building Intelligent .NET Applications to temporarily satiate the need to expand the applied library of knowledge for all things .NET.
Mike Riley
Rating:
Title: Building Intelligent .NET Applications: Agents, Data Mining, Rule-Based Systems, and Speech Processing
Author: Sara Morgan Rea
Publisher: Addison-Wesley
ISBN: 0-321-24626-8
Web Site: http://www.awprofessional.com/title/0321246268
Price: US$44.99
Page Count: 312 pages
