Mark Watson is an accomplished programmer and writer of thirteen books on various technical topics. An expert in artificial intelligence and language processing, Watson has advised the Defense Advanced Research Projects Agency and is currently developing KnowledgeBooks, an information management tool. Recently, Watson released two books, Practical Artificial Intelligence Programming in Java and Loving Lisp, under the Creative Commons Attribution, No Derivatives, Non-Commercial license. He is drafting a third book, on software design, under the same license.
Watson spoke with us recently about his motivations for licensing his works, what he has gained from the experience, and his thoughts on the Semantic Web.
Creative Commons: What motivated you to publish under a Creative Commons license?
Mark Watson: I liked the idea of using a standard license. I also thought that it was a way to support the good side in copyright, people like Professor Lessig. I feel that the recent extension of copyright for longer terms is not in the public interest. To be blunt, I feel like a few large corporations are buying off Congress, and I don’t like this trend.
I still do write non-free published books, but I also have a desire to give something back, and writing free web books under a Creative Commons license fills that need.
CC: What experiences have made “giving back” important to you?
MW: I receive requests from teachers in non-industrialized countries for permission to reprint part of my published books for their classes. Clearing the rights in these cases is, in general, too costly for my publishers and for people asking permission. I travel a fair amount, and I feel that our world is a small place and we should try to get along. We in the U.S. have got a free ride on many things, and we should seriously count our blessings.
I value the times people have made use of my work. About five years ago, I spoke with a computer science professor in Peru who had copied just a chapter out of one of my books for his class. A few years later, we ended up working together briefly at Intelligenesis [an artificial intelligence corporation]. Small world!
I believe in a gift economy. I believe in a take-what-you-need-and-leave-some-for-others philosophy.
CC: To fund your writing, you offer advertising space in your books and ask your readers for donations. Have your efforts been successful? Has your Creative Commons licensing of these books attracted people them?
MW: I have never sold any advertisements, but I do receive about $40 per month in donations. I like this for two reasons. First, if someone takes the time to send me a few dollars via PayPal, I take that as a compliment. Second, the donations do pay for my bandwidth.
The advertisements that you see in the web books are for a few of my favorite charities; I hope that they get some money as a result of those free advertisements.
CC: Artificial intelligence and natural language processing—in the same spirit as Creative Commons’ RDF metadata—may be helpful in fulfilling the vision of a Semantic Web. Tell us a little about how these tools will do this.
MW: The Semantic Web is a version of the Web that you can communicate with like a person. Using software that can accurately process and understand the information on Web sites, you will be able to ask search engines to retrieve specific facts, rather than pages containing some list of words.
Imagine, for example, that you want to research a company’s financial history before buying stock in it. Rather than telling a search engine to look for key words like “stock” and the company’s name, you could ask the search engine, specifically, who’s on the board of directors, how the stock has been doing, and what the SEC thinks of the company.
The Semantic Web is a bit of a dream at this point in time. There are two interesting aspects to the Semantic Web. First, we need to convince human authors of webpages to take the time to add RDF tags to help identify content, which Creative Commons encourages. Second, we need automatic artificial intelligence and language processing tools that can categorize text, and extract keywords and phrases.
My KnowledgeBooks demo is one example of how this can work. When I spider news sites—I have written permission to do this—I detect human names, place names, Reuters news categories, and key word phrases. I have also tried to build a system that searches by concept.
It is very difficult, in the general sense, to write software to pull meaningful information from arbitrary web sites. Having people help search engines out by putting RDF tags on their sites is a good start.
If and when the Semantic Web becomes more of a reality, I think that we can expect software agents that will enable us to find information that we need quickly. Google is a great tool, but it is still just a keyword search—still just a start.