Tim Berners-Lee Says the Time for the Semantic Web is Now
In an hour long interview posted today about the Semantic Web, W3C Director Tim Berners-Lee says all the pieces are in place to move full steam ahead and realize the potential of a world of structured, machine readable data. Available as a part of the Talking with Talis semantic web podcast series, the interview (listen here) is summarized on interviewer Paul Miller's new ZDNet blog dedicated to the semantic web. A full transcript is available here.
It's an important conversation and a good introduction to what the semantic web is. Also notable is the way that Berners-Lee sees Semantics and Data Portability as very related. Some highlights are excerpted below.
My standard explanation of the value of the Semantic Web is this:
Once our software is capable of deriving meaning from web pages it looks at for us, then there's a whole lot of work that will already be done, allowing our human, creative minds to reach new heights.
In the interview with Miller, however, Berners-Lee emphasized that it's not just about web pages. He told Miller that that the core pieces are in place today for developers to build robust Semantic Web applications;
“I think… we’ve got all the pieces to be able to go ahead and do pretty much everything… [Y]ou should be able to implement a huge amount of the dream, we should be able to get huge benefits from interoperability using what we’ve got. So, people are realizing it’s time to just go do it.”
On the topic of challenges still faced, Berners-Lee said:
“There’s an awful lot of data out there. And I think, one of the huge misunderstandings about the Semantic Web is, ‘oh, the Semantic Web is going to involve us all going to our HTML pages and marking them up to put semantics in them.’ Now, there’s an important thread there, but to my mind, it’s actually a very minor part of it. Because I’m not going to hold my breath while other people put semantics in by hand… So, where is the data going to come from? It’s already there. It’s in databases…”
Other topics of the interview include whether leading social networks are likely to implement semantic web technologies, how semweb engagement benefits companies and what users can do to move the technology forward.
We've cover the Semantic Web extensively here at RWW. See below for a list of posts on the topic.
Share
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
Thank you for including the link to the transcript ( shouldn't it have been buried in the mark-up, requiring a "View Source", given the subject? ).
I wish more people would include the transcribed text of podcast they make or reference? IS that expensive to do or something?
Is it just me ? but I still don't understand this whole semantic web stuff. Data cloud I can understand, api's , mashups.. ok thats a good idea, but so far the best example I can say is a site, using maps for something. Other than that nothing impressed me that much. Semantic web, seems like opening up these api's and doing something with all this data, what.. i don't know.. thats what berners-lee is for.
This guy does one little thing in his life like invent the Internet and now people will forever listen to him.
Although in fairness I guess he didn't get as much traction as Al Gore did for doing the same thing.
@confused "The Semantic Web" means different things to different people, and please consult the blog post authors and the majority of commentors of this blog for more info - they are far smarter than I.
To me the promise of the Semantic Web is simply Freedom. That's with a captiol "F" like actor Mel Gibson sreamed out at the end of Braveheart.
All the world's data, be it brilliant white papers about cancer cures or LOLCATS, is currently trapped in fortress. These fortresses make you sign -up for an account at best, pay for the "privilege" to access the data at worst.
An open, read write network of the world's data would let us all get thing done, and not worry about whether or not I have the latest version of .NET installed locally on my computer. We have a duty to free the world's knowledge and let future generations look back and say "thanks" ( As opposed to "WTF is an ActiveX plug-in and why do I have to have it in order to read this historical document!?!?" )
Again, take my childish interpretation for what its worth ( $0.02 ) and read more at Wikipedia or right here on this blog.
I gotta plead ignorance on this as well. :( It seems like specialized markup, but remember, browsers can't even agree on html/css markup. Now we're going to surrender our data to this mess as well? It reminds me of the hype surround java years ago. That was supposed to solve all our programming/platform woes...we'll see if this semantic web gets us any closer.
My attempt to explain semantic technologies...
Traditional computer technologies focus on processing data. Semantic Web technologies focus on processing metadata. The metadata is processed using ontologies. Ontologies define the relationships, the rules and the logic that apply to the metadata.
Using Semantic Web technologies, metadata is given a well-defined meaning enabling computers to process higher-level “knowledge” about the data.
For example, the statement "John is the son of William" if defined in a genealogical ontology would allow the system to infer an additional relationship - William is the father of John.
For those having trouble with the SemWeb concept, visualize the world's databases turned inside out and openly exposed for others to peruse, extend and remix. In a sense, its somewhat like the hooks and loops of a loosely coupled, semantic velcro.
Does SPARQL still lack aggregate function specifications like "count" and "group by?"
Counting results in a query language is basic functionality. Imagine if SQL was lacking a count operator, you'd never be able to count the total rows given some criteria.
Yes kids, it's that time again :).
i strongly agree with your "reach new heights" thinking...
no matter how fine the mesh, human consciousness will be more subtle, and having a lot of the grunt work conceptualization taken care of machine-wise will reallyt allow this to become obvious...
the genome guys thought for a few minutes they got to the fundamental building blocks, and found that, no way, there was so much more at subtler levels that it only increased the questions...
same with the semantic web thing...
the danger? algorithms taken as reality, in the same way your insurance sx canceled by a computer because of some obscure programmer's decision
anyway, it is going to arrive sooner rather than later, after all, technology is just the out-picturing on the material plane of the natural and inherent abilities of consciousness, and we haven't even gotten to first base
enjoy, gregory
Looks like the SemWeb folks are even more confused than ever. An article blob in a database indexed by time doesn't make it any more useful than a simple web page. The meta tags are proven ineffective. None of the SemWeb techniques is scalable, which is the Achilles heel of the current incarnation of SemWeb.
To achieve the real goal of semantic web, you need to practically solve AI. The current NLP and ML techniques are very limited and not scalable, as language model building is still pretty much an art. We'll probably see some very primitive SemWeb stuff in English (forget about other languages) soon. It'll still be unimpressive and not that useful, compared with just full text web search from Google etc.
The prime time for semantic web is "now" + 5 years, with "now" a moving variable.
If so many of us are having trouble figuring out the exact meaning of an idea, my guess is that there is no idea out there!
Where are the applications? Because without applications the Semantic Web might as well be "Write Only Data".
Posted by: Julian Bond
|
February 28, 2008 2:39 AM
The Semantic Web activity is both broad and diffuse, spread across universities, companies and individuals around the world.
I don't think it's 'confused', so much as demonstrating the broad range of purposes to which semantic technologies can be put. You can talk to lots of people, all of whom see benefits in semantic technology. But there are so many potential benefits that you'll tend to just hear about their own (sometimes narrow) area of interest rather than seeing the whole picture.
Alex Iskold, for example, is using semantic technologies in his products over at AdaptiveBlue. Those products are very different from our Talis Platform ( www.talis.com/platform/ ), so his corporate view of SemWeb advantages would be very different from ours. It doesn't mean he's wrong and we're right, or vice versa.
There are a growing number of companies investing heavily (Reuters with Clearforest for one) in deploying semantic technology, whether in whole new applications like Radar Networks' Twine or (perhaps more likely?) behind the scenes in longer term enhancements to mission critical core infrastructure.
With the exception of consumer offerings such as Twine, most end users will probably never 'see' the Semantic Web, and nor should they. They applications they use will 'simply' get smarter.
For the people building those applications, infrastructure such as that we're investing in with the Talis Platform is intended to make it easier to get started, and easier to gain the benefits of the Semantic Web without having to go further into its arcana than they might wish.
Just like is said in the article, tagging content in a web page with tags is the smallest part. (he said something like he's "not going to hold his breath until it's done). The reason being that computers can scree-scrape and guess tags for data (my guess is that this is what Twine has to do since most sites have minimal specialized tags).
In general, the magic happens with the Ontological relationships defined between these tags (which can be an immense amount of data), and the different triggers and responses that can be programmed in algorithms according to such ontological relationships.
I'm no Semantic expert yet, and I'm only here to learn more, but my guess is that eventually robust Semantic applications will be able to be plugged into play by software developers. They'll be able to plug in a particular Ontological tree set and plug in a particular Natural Language Processing algorithm set and connect the two. The natural language processing algorithm would essentially connect different actions users use to the words used in the ontological tree, ultimately delivering responses to, say, a user in a social network based on HIS actions.
If there are some experts out there, I'd love it if they "correct me if I'm wrong."
James
from
FaceySpacey.com, Your One Stop Social Media Shop
A way to find printer cartridges that are compatible with my printer...? Wow, that is clever. All that money (we are talking millions and millions) mainly from the tax payer to fund researchers and companies for this? Let us not forget all the trees lost so that they can print the volumes of papers and books on this subject.
If it can be of any help for people wanting to know more about the semantic web, here is a little web site we are recently putting online : www.websemantique.ch
One of the purpose of this site is to explain simply what the semantic web is, and how it can be put into practice.