Monday, January 15, 2007

Paper Review: two Websense columns by Danny Ayers

Recently, Danny Ayers published two successive Websense columns on Internet Computing: "The Shortest Path to the Future Web" and "From Here to There." In these two columns, Danny presented his views of the current web moving towards the Semantic Web. This is a brief summary of the two articles based on my reading.

In his first column, Danny discussed a path to the future web, which, by his words, may or may not be the shortest one, however. Semantic Web technologies offer a logical model that are full of potentials. This is what Danny want to deliver to readers.

This first column starts with a summary of the traditional web. Danny emphasized two points: simplicity of the web and communication model. The majority of web users do not have much knowledge of web technologies. Simplicity is surely a key of success. Moreover, the strength of WWW lies on its communication model. Most of the time, WWW is exciting because it links everybody. By this sense, the evolution of WWW very much depends on the advancement of its communication model.

Then Danny discussed the ways of revising the web. I am particularly interested in his comparison between incremental and leaping progresses. In fact, Web 2.0 shows the combination of these two types of development. With several years of incremental progesses on its key techniques (such as tagging and blogging), the web jumps to a new level (becoming Web 2.0). The aggregation of many minor increments can eventually lead to a fundamental leap. Is the same process happening right now on the realization of Semantic Web? The current Web 2.0 hype can bring Semantic Web researchers many hints. I believe this is what Danny want to express.

In his second column article, Danny quoted an interesting proverb by George E.P. Box: "All models are wrong. Some are useful." This is a very interesting observation, though I am still not very sure how it is applied in the article.

Semantic Web is thought to be the web of data. I have seen this representation in many places. But I am a little bit afraid whether we have pushed this argument too far. Semantic Web adds machine-processible semantics for web data. But is this process of semantic annotation objectively or subjectively? The statement (web of data) likely emphasizes more on the objective side than the subjective side. As a result, the role of humans on deciding the meanings of data becomes unclear. Will it be a big problem later?

Are we on a transitional web now? If yes, are we within the transition from Web 1.0 to Web 2.0, or within the transition from Web 2.0 to Semantic Web? I think these are interesting questions. Microformats have been mentioned by more and more researchers and developers. Will micorformats be a standard presentation in future web? Or is it itself also a "transitional term"? We may still need time to answer these questions.


Danny said...

Thanks Yihong!!

Yihong Ding said...

Thank you, Danny. You have done great work.