Monday, June 25, 2007

Article Review: Embracing "Web 3.0"

Ora Lassila From time to time, I wrote short reviews about some interesting articles. To be honest, being a critic is much easier than being a creative writer. All these orginal authors are tremendous. They addressed excellent insights on these articles. My purpose is only to share my understanding after reading their work, and try to broadcast these brilliant work to more people.

James Hendler Not long ago, two leading scientists of web research, Ora Lassila and James Hendler, published an article titled Embracing "Web 3.0" at IEEE Internet Computing. This is a great article addressed the understanding of the buzzword "Web 3.0" and even the web beyond it from top Semantic Web scientists. This article begins with the addressing of a popular New York Times article by John Markoff, which claimed Web 3.0 to be the Semantic Web. Many Semantic Web researchers, including these two authors and me, do not like this naming. But all of us agree, as addressed in this article, that "we enthusiastically embrace the technologies it is bringing to the field."

Brief Summary

This article contains three parts. At the first part, the authors briefly introduced some background of Semantic Web probably to the readers that are lack of knowledge about the Semantic Web. At the second part, the authors presented their understanding of what "Web 3.0" might look like. In particular, the authors have addressed the adoption of RDF and SPARQL in the industry domain. At the third part, the authors addressed their dream beyond Web 3.0. As a matter of fact, the descriptions seem going back to the original dream of Semantic Web, which have been discussed repeatedly among Semantic Web researchers for years. In the authors' mind, the so-called "Web 3.0" is nothing but a short period of infant Semantic Web.

Comment and Thought

In this article, the authors presented that Semantic Web "supports ... a view of information processing that emphasizes information rather than processing." Despite of agreeing with what the authors wanted to emphasize, this sentence itself is dangerous. This sentence delivers a hint that Semantic Web is primarily a stationary web (since it is more about information) but not an active web (since it is less about processing information). This latent message becomes very dangerous because without the ability of processing, what on earth can we expect the production of semantics? Information is produced for processing. But without the focusing of processing, the momentum of producing information loses. In fact, this is the problem I repeatedly blogged. This is why the realization of Semantic Web becomes so slow, and it is still quite overlooked by many people. Unless we start to refocus the attentions onto the processing side, Semantic Web would not be realized in near future.

A lesson we have learned from the previous "AI hype" is that "you can't sell a stand-alone 'AI-application'," addressed by the authors. I totally agree with this point. Unfortunately, however, it seems that many people still do not learn from it. This lesson basically tells us that semantic Google is impossible to be built. No company by itself can build a super-program, even if it might be very well paralleled and distributed, that semantically processing all knowledge on the web. In order to realize the real, pragmatic semantic search, collaboration of almost every web user on the web is a demanding. That is, the solution could only be a search network that is operated by every one, but NOT operated only by one big node. This is thus the vision of the semantic search web.

"Web 2.0 is most a social revolution in the use of Web technologies," addressed by the authors. Nevertheless Web 2.0 is a social revolution, it is far beyond a social revolution. Web 2.0 is also a major landmark on web revolution. Unlike many other Semantic Web dedicators, I am kind of like the term "Web 2.0." From the view of web evolution, this term very well reveals a fact that our web is evolving and it is now at its second major stage of its evolution. The hype of Web 2.0 changes not only our real human society, but more importantly it also changes our virtual human society, i.e., the World Wide Web itself. The quality of web resources has been significantly upgraded. Semantic Web researchers must not only focus on Floksonimies and Microformats. Thought these two things are unquestionably important, other things such as the implementation of community collaboration might be at least as important as, if not more important than, the mentioned two technologies.

Web 3.0 is more than RDF and SPARQL. In short, the trend of web evolution is the web being more and more alive. Machines (or machine agents) on the web will start to think of things by themselves based on their learned semantics. No single machine could think all the knowledge. But only if individual ones are thinking of their limited knowledge, the collaboration of all of them can handle close to the entire set of web knowledge. Similarly, no individual humans know much of knowledge. But the whole set of human knowledge becomes so huge and it is so great because we are not just individuals. In contrast, we are collaborated people and our society has been evolved to well cooperate individual knowledge. This is what Web 3.0 and the ones beyond it would deliver to us.

1 comment:

Anonymous said...

I loved your blog. Thank you.