Foreword

When the public talks about Artificial Intelligence, in most cases they actually think it still belongs to the science-fiction realm. The great Alan Turing predicted in 1950 that in about fifty years, "an average interrogator, after five minutes of questioning, will not have more than a 70% chance of identifying whether or not a computer is indeed a computer." However, the enthusiasm for a real "thinking machine" has slowly fizzled, and we are now taking incremental, evolutionary steps toward "smarter computers." Researchers are now tackling smaller problems than the thinking machine, but it doesn't mean that AI research hasn't produced real-world results. Neural networks technology has been used for a wide variety of tasks, from stock market prediction to computer games; speech recognition is now entering mainstream; various machine translation services are available online; handwriting recognition, as well as optical character recognition, can now translate arbitrary (type) written script into text.

Being involved in various AI-related development projects for almost 15 years, I did my share of programming in Prolog, LISP, Smalltalk, C, Java, and similar languages and development environments. Judging from my previous experience, the .NET framework and ASP.NET represent an exciting and important evolutionary step in the software development world. They provide a feature-rich application and development environment and are language and platform neutral although more efforts will have to be invested to achive full platform independency through projects such as Mono. It is all about reducing the burden and the complexity of software creation. It is also about winning the hearts of developers all over the world through the support for collaborative projects, workspaces, and even the open-source movement, something that Microsoft has criticized strongly at times in the past. .NET does require a shift in developer mentality, but the time invested in it quickly pays off.

So it is time to put these two, Artificial Intelligence and .NET, together. Both technologies are innovative; both are ambitious. Considering the amount of hype surrounding almost all AI-related technologies, it should be easy to find your way through, right? Think again: .NET is a newcomer in this area, and don't expect to find plenty of resources on AI development using its tools and languages. Most code samples and tutorials are written using traditional AI tools, and converting them to .NET can be a slow and painstaking process. This is where Sara Rea fills a huge gap with this book and its practical, hands-on approach. The four AI-related areas she covered offer unlimited potential for business-related applications. After all, a typical professional software developer doesn't have time to reinvent the wheel and needs a focused and pragamatic guide to this vast field. From speech applications to data mining, rule-based systems and agents, this book offers a solid instructional text of basic theory, the principles from which it derives, and how it is practically applied to develop advanced products.

As I was witnessing this book being written, I quickly discovered Sara's passion for knowledge and a talent in making hard things easy. I'm sure you'll appreciate that many advanced topics are presented in an easy-to-digest form thanks to this talent. Unlike many other technical books, it will not leave you feeling overwhelmed, but motivated to apply the techniques described inside to the real-world problems. And that's what software development is all about: building better and smarter solutions. Only this time, they'll be really smart.

Denis Susac

CEO of Mono Software,
http://www.mono-software.com