Tuesday, December 25, 2007

Happy Holidays and Seasons Greetings

Seasons Greetings to all. It is indeed a wonderful holidays as the Google Scholar has published an important piece to the Semantic Web literature. He's done it again, writing an concise and cogent piece on the key elements which differentiates Web 3.0 from Web 2.0. In other news, a reader recently made a comment from a previous entry which I found to be very interesting. Here's what he said:

I (as a librarian) found the article and the whole topic very important. I especially enjoyed the conclusion. You wrote that "Web 3.0 is about bringing the miscellaneous back together meaningfully after it's been fragmented into a billion pieces."I was wondering if in your opinion this means that the semantic web may turn a folksonomy into some kind of structured taxonomy. We all know the advantages and disadvantages of a folksonomy. Is it possible for web 3.0 to minimize those disadvantages and maybe even make good use out of them?

My response? It'll sound cliched and tired: it's really too early to tell. But although it's murky as to what the Semantic Web will look like, all directions point to the possibility that folksonomies will play a key role. Here's why:

(1) Underneath the messiness of the Web, is a fairly organized latent structure, whose backbones are web threads. A scale-free network is significantly dominated by few highly connected hubs.

(2) What this means is that folksonomies and tagging are in fact controlled vocabularies in their own right. Lots have been written about this. Recent studies have shown that the frequency distribution of tags in folksonomies tends to stabilize into power-law distributions. When a substantial number of users tag content for a long period of time, stable tags start appearing in the resulting folksonomy.

(3) Such a use of folksonomies could help overcome some of the inherent difficulties in ontology construction, thus potentially bridging Web 2.0 and the Semantic Web. By using folksonomies' collective categorization scheme as an initial knowledge base for constructing ontologies, the ontology author could then use the tagging distribution's most common tags as concepts, relations, or instances. Folksonomies do not a Semantic Web make -- but it's a good start.

No comments: