Now let me speak to Thomas Paul's question about Semantic Web applications. It is not so easy to answer in a short space because many people have different ideas of the scope of the Semantic Web. Also, it depends on whether you are interested in the more futuristic (or optimistic) kinds of applications or something that could be done now or in the near future.
My book does have some application scenarios, and they are available on line in the first chapter, which can be downloaded from the publisher's web page for the book (see
www.manning.com). That chapter also sketches out the range of notions of what the Semantic Web is supposed to be. Near the end of the book, I revisit the scenarios with a detailed discussion of each part of them in light of all the other material in the book.
To give some of the flavor, I will summarize parts of one of the book's scenarios. Suppose you are at a conference, and you meet a participant, Sam, who gives you his vcard information - like the vcard data that some people already attach to their emails - name, employer, email, etc. In the scenario, this data is automatically ingested by your semantic web-enhanced PDA.
At the end of the day, you review the day's events, and find that your PDA has found that Sam seems to be a good candidate for a job opening an another department at your company. It even offers to notify the right person back at the company. During the day, it seems, your PDA went to work and checked out everything it could find about Sam, and found a match between his background and the job requirements listed by your company.
You want to know why the PDA is making the suggestion, and it has a button to press asking for an explanation of the reasoning. It has another for more information. After you digest the explanations, you decide to go ahead and have the PDA give the information the the appropriate person (which it already knows).
There is much more to the book's scenarios, but this should give the idea. Ideally, there would be a fairly seamless integration of various ontologies (to allow the PDA the work with the terms it found), workflows (so the PDA would know it should scan for likely candidates, perhaps), discovery of semantically useful data, and reasoning to draw sensible conclusions and to be able to explain them.
That's the kind of picture that Semantic Web proponents have in mind.
Does this answer your question well enough? If not, please post back so I can know what the best direction to go.