|
|
2 of 2 people found the following review helpful:
3.0 out of 5 stars
Needs more details, June 19, 2003
This book outlines a proposal to integrate artificial intelligence and wireless technology into the World Wide Web in order to make it more powerful and more tuned to the real needs of the user. It is an interesting proposal, but omits discussion of some important issues. The book is targeted to an audience of developers, engineers, researchers, and IT managers who need to understand how to deliver, via wireless technology, intelligent processes and services. Chapter 1 is an overview of the five areas that the authors feel are needed to form the Intelligent Wireless Web. All of these are viable and desired from a technical standpoint. However, from a human factors standpoint, one of them is somewhat troubling, namely the need for having voice activation for the user interface. This could be extremely annoying if one is working in the now popular cubicle environments, due to the noise level generated from user's speech. Privacy issues could arise too, since voice patterns are easily recorded. Making the transition from dumb/static Web applications to intelligent/dynamic ones is sorely needed, but voice activation/recognition should be the problem of those who are working in other areas of machine intelligence, such as robotics. Of course, if work environments evolve into more private scenarios, the author's proposals for voice activation could become viable. Chapters 2 and 8 concern speech recognition. I did not read these chapters so their review will be omitted. In chapter 3, the authors discuss how wireless technology could be integrated into peronal area networks (WPANs). The authors here exhibit a keen awareness both of the technology and the human factors involved in creating what they call a "Personal Space". Home automation will be slow-going perhaps at first, due to legacy systems now in place, but it is highly desirable from the standpoint of energy conservation and home security. To prevent government and other forms of malicious intrusion, wireless security will have to be top priority before the Intelligent Wireless Web is implemented. Chapter 4 is an overview of the basics behind both wired and wireless networks, with the goal of merging them effectively. The authors are clearly advocating the use of LMDS for high-speed wireless access. However, they do not discuss any performance studies to give more weight to their arguments for LMDS. "Project Oxygen" is discussed as an approach to accomodate mobile and stationary devices, and for moving away from TCP as a congestion manager, but the discussion is too brief to be helpful. In chapter 5, the authors discuss the status of mobile wireless, IP version 6, and Mobile IP. The authors are a little more quantitative in this chapter, mentioning for example the inability of TDMA to deal with bursty data flows, but no details are given. A fairly detailed overview of "third-generation" mobile wireless technologies is given however. Performance issues are not discussed though, and it would have been interesting if the authors would have included a discussion of MANET. Chapter 6 is a general overview of artificial intelligence and how it might be applied to Web protocols. As in all discussions on AI, controversies and disagreements will arise in the mind of the reader, but the authors are fair in representing the main ideas, considering the relatively short length of the chapter. The discussion on distributed AI is the most relevant for the book. In chapter 7 the authors continue the discussion on AI with the goal of seeing to what extent it can be incorporated into the Web. I was glad to see a discussion of the Cyc application in this chapter, even though it was very short. From the author's standpoint the Web currently does not really express intelligence, since it does not adapt, a necessary requirement for learning. A "learning algorithm" is defined as a process that extracts data from a database to serve as its input, and then performs a set of operations on the input, giving finally an output that represents learning. The authors feel that the Semantic Web holds much promise for building an intelligent Web, and outline several tools, such as XML and RDF, that assist in the construction of the Semantic Web. Particularly interesting is the discussion of the need for self-organization in order for the Web to be considered intelligent. The property of self-organization will also be the most problematic to implement, due to the extreme distrust that some now feel against software that has not been validated by a human. This is especially the case for those having to deal with medical records and information on human health. So why even attempt to build the Intelligent Wireless Web? The authors attempt to answer this question in Chapter 9. They conclude, based on Moore's law, that wireless chip technology will allow cellular carriers to build networks for less than $100 per customer. They never however answer how much intelligent applications over the wireless Web will improve productivity. This can be accomplished to a large degree with simulation and mathematical modeling, but the authors do not do so. Chapter 10 is an overview of the actual progress in developing the Intelligent Wireless Web. The challenges are considerable, not only from a technical standpoint in the creation of intelligent applications, but also because of legacy issues. The authors are aware of this and give a network schematic outlining an integrated wired/wireless network. Their concept of an Intelligent Wireless Web is a good one, but their justification for it, especially for the use of speech recognition, is somewhat weak. They need to perform a lot more modeling studies to see just how these smart applications are going to behave on the Web.
|
