Challenging the Ideas Behind the Semantic Web

mytrip writes to tell us that after a recent presentation to the American Association for Artificial Intelligence (AAAI) Tim Berners-Lee was challenged by fellow Google exec Peter Norvig citing some of the many problems behind the Semantic Web. From the article: "'What I get a lot is: "Why are you against the Semantic Web?" I am not against the Semantic Web. But from Google's point of view, there are a few things you need to overcome, incompetence being the first,' Norvig said. Norvig clarified that it was not Berners-Lee or his group that he was referring to as incompetent, but the general user."

Re:Googlebombing

[QuantumFTL]

Google does not extract any semantics from content. It merely analyses the linking between websites and connects that with keywords. No semantics here.

I believe you are referring to PageRank, which is one of many algorithms used by google to determine search relevance. This article [seobook.com] discusses their use of Latent Semantic Indexing [wikipedia.org], which is a somewhat crude but effective form of sematic inference which is widely used in the field of NLP.

Sem Web, meet Chicken & Egg

[AlXtreme]

The semantic web is, in my eyes, a typical chicken & egg problem. You've got loads of content on one side, yet current search engines work well enough to not worry about representing that content in a structured way in a markup language like OWL. On the other side, you've got embarassingly few semantic web applications that use structured content. How is a typical web developer going to justify structuring the content on his side if he can't point to an example how it could improve shareholder value? What would exporting our databases in OWL currently solve?

True, the web had a similar problem, however creating a webpage is a lot more interesting (you see the results directly, how terrible they might be you do see a result) than structuring data. The latter takes a lot more work, and the direct benefit just isn't there.

Sem-Web-like standards like RSS, XML and SOAP have become mainstream, but primarily because they fill a gap. The adoption of RDF or OWL simply doesn't solve anything. Yet. It would be cool to let agents loose onto the semantic web and retrieve them together with a summary on a certain subject using a multitude of sources, but as long as it's easier to Google I don't think it would generate any interest outside academia.

Feel free to prove me wrong though.

Re:Always bet on the million monkeys

[mgblst]

People who want to add extra information to there page can, it doesn't all have to change at once. The people who do add semantic information to there page, will be indexed better, or by a different browser which producers only relevant results - this is a huge advantage. Then, this will be popular, and more and more people will add the extra information (which for sure, takes extra time). If people spam the system, or put in incorrect information they are excluded. It is possible, and it is the next step - it is a solution to the problem of spamming.

Re:Googlebombing

[navarroj]

> Google has the right idea, automatic extraction of semantics from content.
>
> Google does not extract any semantics from content. It merely analyses the linking between
> websites and connects that with keywords. No semantics here.

Google does extract semantics from content in a few particular domains: addresses and bussines info for Google maps, show times and additional information on movie searches, dates and appointments from Gmail to Google Calendar, ...

The semantic web has already started. Now we only have it in a few and simple enough domains but, I agree, this should be the right way to go.

Re:It's really, really difficult...

[Bogtha]

I still don't think I truly understand how RDF is supposed to work...

I don't think anyone does.

RDF's core idea is simple. Give everything a URI. Express relationships as a set of three URIs, (subject, property, value). So you might have (#me, #friend, #bob) expressing the idea that Bob is a friend of mine. Or you might have (#photo, #contains, #me), expressing the idea that I'm in a photo.

RDF is little more than a mechanism for expressing relationships. It doesn't give software the ability to understand those relationships, you need to build that on top. RDF just helps you solve the relationship problem in a generic way. So, for example, even if you have (#me, #friend, #bob), that's still meaningless until you write software that knows the #friend, #spouse, #employer, #whatever URIs are relationships between people. For instance, you could build a social network like Friendster, only decentralised - and people have, with FOAF [jibbering.com], because they've agreed that particular URIs express particular relationships.

Tutorial on the Semantic Web

[CaptSolo]

Here is a Tutorial on the Semantic Web [w3.org].

Pay attention to the slide #22 which shows how data from different sources can be merged together. This is one of key differences between XML and RDF - to merge XML data from a number of different schemas one would need to create an application that processes data in these schemas and generate merged data (possibly inventing a new schema to represent the merged information).

In RDF that happens "magically" - in order to merge heterogenous data you don't need to do *anything* - just put all the information in an RDF store and it merges. If the data to be merged change no modifications to the store are necessary - it is like a bag that can hold anything.

Re:Damn

[salmon_austin]

In the U.S. regular cage fights are 3 rounds, with championship fights being 5 rounds. There are no 10 round cage fights that I am aware of anywhere in the world.