Sunday, May 20, 2007

The semantic Web is dead?

Mor Naaman from Yahoo Research Berkeley has a thoughtful piece on the state of the semantic Web. In a nutshell he declares:
The Semantic Web is not dead; but the grand vision of “A Semantic Web” will not be achieved, despite Tim’s noble efforts. Instead, we will see the Emerging-Semantics Web, derived from how people/developers use lightweight formats and tags on popular platforms such as Flickr and YouTube.

Where he mentioned a bunch of reasons for why this is going to happen it mostly comes down to this:

There is no way that we can engage the masses in annotating media with “semantic” labels. At best, we can get the people to annotate content (such as Flickr images or YouTube videos) with short text descriptions or tags.

I've been following this stuff for a long time, and semantic Web technologies are years away (if ever) from providing "average" people with something they will actually use. Things aren't going to stand still waiting on the research community to catch-up. The Web is already moving further away from the vision of a semantic Web. We are seeing some walled-gardens (islands) of semantic content emerge, but these are not efforts that will touch the masses.

If you look at user behavior you can see the future, and it is one of compromised and scaled-down functionality rather than more. Less scales. You see this being played-out in the discussion of Google's new universal search. The question is being asked, "Google Universal Search - is Vertical Search Space Finished?" The answer to that question is a resounding yes! Is vertical search better? Yes! Does it matter if it's better? No! We've always been able to build better search of our own unique databases using SQL queries. So what? People won't come to your site to use them. Just like they won't use advanced search options when they are just one click away on Google's homepage.

In the end, all that matters is what people are willing to do. So, I'm with Mor, in declaring the semantic Web as it has been envisioned, dead. All of these so-called "advanced" functions are dead except in very controlled user environments, and no one would ever be foolish enough to describe the Web as controlled.

UPDATE: You might like taking this poll on universal vs vertical search. Not to bias you, but I lean more towards "dead meat" than "goneburger".


Floyd Davenport said...


Great post. If there is one thing we have learned over the last few years, communities will decide where to take technology and sometimes it is a complete surprise.

I like how you summed it up "it only matters what people are willing to do". We should post this up on the wall of every office.

I think it also speaks to your earlier question on adoption of technology.

Kevin Gamble said...

I wasn't too sure about this post so I appreciate your feedback.

I hear a lot about how we *have* to provide a mechanism to customize/localize content. More and more I'm thinking, if it's local write it that way or link to something that is more specific, but don't try to do this with technology. First, off you bury it behind some logic and no search engine is ever going to decipher the nuance of the data.

So I'm totally with you--keep it simple. Don't try to fix what isn't broken, and on and on an on...