Yahoo Goes All Semantic On Us

San Francisco: Arrington thinks that Yahoo's announced support for the Semantic Web standards -- hCard and a bunch of other microformats standards, and other competitors' efforts (Amazon OpenSearch) -- is a really big deal.

[from Yahoo Embraces The Semantic Web - Expect The Internet To Organize Itself In A Hurry bu Michael Arrington]

What does all this mean? It means we can expect the web to get itself organized, in a hurry. At stake is a significant amount of traffic from Yahoo search, and anyone else that may choose to build applications on top of this data.

But is it?

Yahoo is certainly trying to shift the search battle away from Google dominance into an area where they at least have a fighting chance. And there are various areas where great advances are likely.

However, I maintain that formats embedded in the pages won't catch on the way that social gestures -- links, clicks, bookmarking, comments, and other references that casual and not-so-casual users leave behind -- will always be more interesting than embedded ontological information. Rather than authors and publishers declaring what sort of information they are embedding in their web pages, users could operationally indicate what they think they see.

I am more interested in unstructured, bottom-up information than top-down, structured information. So, ultimately, I don't know about this, at least not as a first-order phenomenon, involving sites like Linkedin and Facebook.

At the same time, the information that user festoon the Web with could be structured using these standards, and I hope they will be.

And certainly Yahoo, Google and others will exploit that information to the extent that it is interesting to do so.

But it I wonder how compelling it will be in the long run.

Comments

[phil jones]

+1

The whole thing is crazy ... what *might* just help is if tool makers (eg. sites) added really easy XML markup tools to their tools. Maybe Orkut / Facebook widgets of the "define yourself" kind could make data available in microformats ... but there's so little incentive for anyone to mark their data up to begin with (and too few disincentives to lie) that the usefulness of the search engines will never become apparent.

[Todd Sieling]

I think Yahoo has a better shot at this as well. Yahoo has been pretty warm to the idea that search needs a human grooming or input component ever since their 'catalog of the web' days. Google, on the other hand, has always trusted the algorithmic approach the most. Both can adapt and change, of course, but Yahoo's DNA seems closer to an approach that uses the human/gestural component.