Sting
INSERT: 1 Richard Smith
"I think if I was suddenly shot forward 30 years to this conference, or whatever follows it, I think we would look back and think what was going on in 2006 was pretty primitive stuff!"
Sting
INSERT: 2 Tim Berners-Lee
"In a way, as I said, we are really at the beginning."
Sting
INSERT: 3 Patrick Sheehan
"There's a maturity of infrastructure which allows people to create business is rapidly and relatively cheaply; but there is a delightful immaturity of business model which is where some of the opportunity is really.."
Sting
INSERT: 4 Carol Goble
"And the academic community came to it Johnny-come-lately really! While they're busy doing some cool technology stuff, 20 million 12-year-old girls were revolutionising the web in "My Space"."
Sting
INSERT: 5 David De Roure
"The only way we're going to join things up is if we have some agreement on how some of these things are going to happen. So having a forum where people can come together to formalise those agreements is absolutely crucial. And the Web conferenceis a really good example of such a forum."
After 15 years we’re right at the beginning; immature business models and teenagers way ahead of the academics!
But this is the forum where the future of the Internet is layout for all to see.
And at the 15th international World Wide Web conference in Edinburgh, more than 1200 delegates got to see that future. They got to understand why we really are at the beginning. The beginning of an Internet radically different from what's gone before.
SIGNATURE TUNE
Hello.
I'm Peter Croasdale.
And over the next four podcasts I aim to give you a flavour of the conference highlights.
In this first one – I'll set the stall out: Try and give you a sense of the key themes and issues that got everone talking.
After that, in the following three. We'll unpack some of those ideas.
And to give you a sense of what those key issues are - here's Tim Berners-Lee - the main man - and officially, Director of the World Wide Web Consortium.
INSERT: 6 Tim Berners-Lee
"I think the spirit of this conference is an explosion in that there are lots of key issues happening we have heard that the Web on mobile phone is very big - we have the the mobile phone initiative. And the semantic Web is very big also - its time is happening. There are a lot of these exponential curves and they're at different places. And there are other things that you will see in the research papers which will start to spread in the next few conferences. Then there are other things which are now becoming very well deployed, like the Web 2.0 which was the document object model presented at conferences in years past. And that has excitement because it's at the deployment stage. I think a lot of people, if I suggested there was one particular thing, there will be an outcry from all people involved in the other things. It certainly is the case that we need a con ference like this because all these things interconnect."
My sense was that at the heart of this interconnection - as Tim mentioned - is the Semantic Web. The growing set of standards that will, apparently, free the data from the depths of those relational database cul-de-sacs.
As far as I could tell, what the semantic web should do is allow computers to share an understanding of what a piece of data is: So that they can do something useful with it. You could, for example, click a travel booking you've made online - and discovering that the computer knows to put that in your electronic dairy - because it has been marked up in a semantic way.
But just to make sure I was on the right track - I asked Nigel Shadbolt, the chair of the Semantic Web plenary discussion, and Professor of Artificial Intelligence at the University of Southampton, for his definition of the Semantic Web.
INSERT: 7 Nigel Shadbolt
"Probably the easiest way to think about it is think about moving from the web we know, which is a web of linked documents, to a web which is a bunch of linked information and data. Now all of our documents, the stuff we used to seeing on the web, contains data and information and we're just very good at getting this information out. And what we are trying to achieve in our standards and methods is to get behind all of the detailed language of the web page and capture the key terms and concepts and relationships that are in there, and represent that in a way that machines can begin to process. So that when your semantic Web server talks to another semantic Web client it can agree on key terms and relations that it's going to be interested in, because you have established beforehand some agreement in what those terms are. And that's one of the key ideas behind the semantic web. It's trying to link not just pages but data and information."
Nigel Shadbolt.
Now, Professor Jim Hendler, was a panelist on the Plenary and one of the co-authors of the original semantic web paper: published in Scientific America, back in 2000.
That set out the idea of number of different layers - the so-called layer cake - which together would enable the Semantic Web.
The first two layers, RDF and OWL are both now W3C standards. With these, the data can be captured and the onotologies, or categories, that explain what the data are, can be described.
But was that enough to get the semantic web out of the realm of the researchers, and into the real world?
INSERT: 8 Jim Hendler
"We are definitely out of the world of scientific papers and research. This week my favourite story to tell you about is Oracle: which has been supporting the RDF model in its current version and will support part of the OWL model in the next version, the semantic Web languages; has just published an article in the Oracle database magazine. So this is the thing that goes to every database administrator who administers Oracle, which is most of them. One of them is the Lilly company, which is one of the big pharmaceutical companies doing drug design and drug discovery. And when I first heard that they were using the semantic web I assumed that that would be at a very low level database for scientist level. But it turns out, in fact, they have many many different databases representing different things happening in their clinical trials. And they're using this to integrate that information for management to know which things are showing which kind of promise on an ongoing rather than at the final report time. If you sort of look at the semantic Web roadmap most of the stuff that is now ready for putting place and use tomorrow is what you might call the start of the roadmap: And there are already things starting to move out of research into small companies and start-ups from the next step. Meanwhile there's a lot of university work to be done on what you might call the top-of-the-layer cake - things that were just beginning to truly understand how to make happen."
Professor Jim Hendler there, of the University of Maryland.
So, the semantic web is out and about. Although, for a concept that has the word semantic - meaning meaning - in the title - you did sometimes wish they'd picked easier words to describe it all.
Maybe that's just me!
Anyway, adapting the ontological construct...sorry...changing subject - the impact of web technologies on science was another major theme.
Professor David De Roure is responsible for Grid and Pervasive computing, also at the University of Southampton.
Now grid computing is basically like an intranet with knobs on. A network of powerful workstations - sometimes even super computers - which share huge datasets between national and international research teams. The pervasive bit, basically means little networked devices anywhere you like - so, PDA's, or mobile phones, or remote sensing equipment.
One of the projects David's looking after at the moment is a flood warning system in the East of England. And it's a great example of a positive feedback loop between grid and pervasive computing. Because they can put these networked, online, sensors almost anywhere they want - along rivers and on the coast - the pervasive bit. You then need serious computing power back at base to crunch the numbers.
INSERT: 9 David De Roure
"Because this is for more information far more often and from far more places, than has previously been used in the models, it is very demanding in terms of the modelling that is used on the back end. So there is some flood simulation going on in the backend which is now requiring an increasing number of PCs to run it. So we have a cluster of PCs doing that. So it is an example of why you need a grid on the backend. But actually we close the loop on that one. Having done the predictions on the water levels in the models in the grid, we then feed that back to the sensor nodes and we can say, "don't bother to report any readings to us unless they are different to these ones that we predict". And that way we can save power, and power is everything in the environment. It is a solar powered project. If you are doing flood monitoring, one of the times you don't have power is when there's going to be a flood. So it is a big issue. And the advantage of closing that loop is that we have this adaptive sampling going on. So we end up with a very complex adaptive system which is a real close coupling of the pervasive world and the grid world -- conceptually. The interesting research question from a computer science and software engineering viewpoint on that is to what extent you can push some of their grid capability back onto the nodes themselves; and that's an ongoing research issue."
As Tim Berner-Lee said right at the beginning – these things are all connected.
So, you won't be surprised to hear that the Semantic Web also plays a part in this project.
Cos of the greatly increased monitoring from this network of sensors the scientists could end up drowning in a flood of unprocessed data. (Sorry - couldn't resist!) But David has the grid working on that - as soon as they take a reading.
INSERT: 10 David De Roure
"You can call it annotation-at-source, metadata-at-source, semantic annotation, call it what you like - but we are capturing all the stuff when it's there. We are not asking people to fill out web pages explaining what their metadata is. We just capturing stuff is going on and on treating it as metadata; treating it as continuous metadata -- as events if you like. And that approach really reveals a lot of the power of the semantic web. And not from the big knowledge management, knowledge engineering ontology end of things, but a more bottom-up view of the semantic Web. Where it is all about capturing this information in machine processable ways; having identifiers that are things that are shared, so that gradually knowledge builds up. As another good pervasive issue, it's a link between the physical world and the world of the web. A lot of people think the semantic Web is metadata, Dublin core, like library catalogues, and what we're saying is -- "no" -- the identifiers relate to things in the physical world. The information we're collecting within the format that you could describe as metadata or you could describe as data now -- are really accumulating knowledge and information about those various physical things. And that's what joins it all together. And that's what gives us the network effect."
And they'll be more on the Semantic Web, Grid computing and research in the next podcast.
Professor Carol Goble from Manchester University was the co-chair for the technical conference. Which mean she helped sort over 700 submissions into the 84 papers that were presented. It also mean she was the best person to ask about what researchers are researching at the moment.
INSERT: 11 Carol Goble
"I would say that the were really two strong themes this year. The first is the "web for people". We've tended to get bogged down very much technologies in the past. For example, the research into the semantic web has been all very orientated around the technologies that we can create for the semantic Web. And we are beginning now to see the issues to do with "well, you know people use the Web - not just machines use the Web". So, examples I would see that this year we have a developing regions track, which is to do with bringing the web to people in other regions who haven't had the massive accessibility that everyone assumes. In fact not everybody has a laptop: not everybody has broadband: not everybody has a PDA. Bringing the web up to people wherever they happen to be. So if they do have a PDA the web comes to them. So pervasive and mobile technologies. The Web being a mechanism for connecting people. So we've seen a big rise in things like "My Space" things like social networks and so on. And the academic community came to it a little bit Johnny-come-lately really! While they were busy doing some cool technical stuff 20 million 12-year-old girls were revolutionising the web it in "My Space". And the Arctic Monkeys were number one by using the web. And I think from the search end, there is this recognition that the search engines are such a powerhouse, but their key technologies is being exploited. So link spamming is actually deflecting the quality of services. So we have to do more work there. So that's the person aspect. And I guess the second theme I would spot is the "people powered web". And what we see is, again, while all the semantic web people were standardising on OWL and the other groovy technologies, the folksonomy, social tagging community, were harnessing the power of people who were just interested in classifying and organising information and being creative with it, to label stuff. So you kind of see this bottom-up rise of the social collective -- a kind of collective intelligence, as opposed to artificial intelligence -- if you will. So I think that we are beginning to recognise that and again in the academic community has said "Hey! That's cool. What can we do, and how can we better understand that and how can we incorporate it into our work; and link with it from the much more academic, top-down, quality, well behaved knowledge management point of view with this rather bottom-up, anarchic, messy, scruffy, collective point of view"."
And you can get more on the messy collective - in podcast three. We'll also have the wining research papers from the conference in that one.
Next!
INSERT: 12 Rohit Khare
"Well it was a very interesting session, where we were lucky to get some people right from the heart of the action, from Google, from Yahoo, Amazon, and from AOL."
Rohit Khare, is the Director of CommerceNet Labs.
And there was a real emphasis on industry involvement at this years conference. Rohit was involved in a number of sessions that looked at the various aspects of new businesses.
This one - aimed at developers - looked at the mashed-up world of web 2.0 and the impact of APIs.
INSERT: 12a Rohit Khare
"The particular thrust we had was - the contribution of that session turned out to be - if you're an entrepreneur and you are to consider starting a business where you wanted to offer web services, the way that these giant companies are, what are some lessons learned about how you can resell services. And the I think even more portentously for the future, if you are a company that starting a new initiative that is going to draw up on those services: I'd like to take photos out of my Flickr account at Yahoo and plot them on the Google map and list the dates that they were taken in a calender and then offer to sell the resulting travel log book on Amazon - this sort of thing may sound a bit fanciful but this kind of remixing is happening as we speak. This is a lot of the Web 2.0 phenomenon. If you're starting a company and you are going to draw upon these services what are the basic business and legal concerns them might go with that. And that was a real fun and unique angle on this conversation was to say that look - what happens when my site goes down because your side went down? What happens when I want users to log into my photo management application are then do they have to give their Yahoo! ID and Yahoo password to get to their account? What happens when I want to combine services from multiple companies and what user names and log-ins would they use then? Or if they upload something that would be considered legal commentary in the application in my country but because someone is viewing it in France or in Germany or in China your other company is now doing some blocking of content or reformatting. All of these are very interesting questions because it's really -- and this is one of the key findings -- the way to think of this is outsourcing. Now, if you think an outsourcing, even without talking about offshoring, just outsourcing a critical business function; like using Amazon's own simple storage service to back up all of your customer data - I would have to ask legally where's my stance on it? Performance wise - what's my guarantee? It's not simply a matter of saying "Hey, they have an API and its $.15 a gigabyte per month that's great!" It's a matter of looking at all these questions but if you are a large company you would have a staff of lawyers and other folks to think about it. When you're making a call and in fact they're only billing you for literally, bit seconds at this $.15 a gigabyte per month rate, you don't stop to think about these things. And that was where - not just the questioning between the panellists - but between the audience got to be quite interesting."
Another thing that intrigued me, was the speed with which business were adopting the Semantic Web. As Jim Hendler said earlier, this is an area that is starting to get some real commercial development time. Chairing a session on this was John Davies, Head of Next Generation Web Research at BT.
INSERT: 13 John Davies
"What was nice about the session was that we had technologists and end-users presenting and talking about what they'd done. We had some quiet advanced technological presentations from suppliers who are very interested in, for example, service orientated architectures, and how semantics are going to help those architectures to scale in the future. But, conversely, we also had presentations from people who were using semantic technologies today. So we had a very interesting example where recruitment intelligence web sites are being automatically crawled and then a database is being populated for use by recruiters. And that was a very interesting example of a scalable application of semantic technology right now. And we heard from the end users of about technology.
Croasdale: Looking into the future now, how do you see the semantic web integrating itself into the business environment? What are the key thing is that it's going to allow businesses to do?
Davies: I think that most businesses actually have a heterogeneous set of data sources often expressed with different database schemas, may be using technology from different vendors. And one of the benefits of semantic technology is that it provides a very lightweight mechanism for wrapping those databases and exposing the data according to a common semantic based information model. There was something very interesting that I noticed Tim Berners-Lee emphasising in the panel this morning, this idea of using semantic technologies for the integrations of legacy data, of heterogeneous legacy systems. And also he was emphasising quite rightly that that's not a heavyweight thing. You don't need to throw away all your old systems - that's the good news. It provides a lightweight technology for integrating those systems and it is of course based on open standards from the W3C, which to some extent will future proof you from forthcoming technologies."
The notion of a pervasive web was the big idea that kicked off the whole conference.
A thought provoking keynote speech by David Brown, the Chairman of Motorola set out his view of the massive economic changes that will take place over the coming years - as a consequence.
Along the way he highlighted the scale of the technical challenges. And, just for good measure, came up with a new name for the mobile phone.
INSERT: 14 David Brown
"Technically and commercially the race to offer customers more personalisable continuity of experience has begun. So the important thing is to put in place a framework around which that increasingly rich environment can develop. That's why open standards and open platforms matter so much. That's why the mobile Internet matters so much. And why it matters that the business model for the mobile Internet is neither the standard IP business model, nor the standard cellular business model. But is instead, a business model based on value rather than volume or time. If today there are more than 30 million web sites, how many more will there be when the mobile Internet adds a couple of billion or so devices, formerly known as mobile phones, into the mix. 10 times more? A hundred times more? A thousand times more? More even than that? Any forecast we may be a rash enough to make today could turn out to be even more conservative than the forecast of mobile phone volumes made by the industry made in the early 1980s we tend to think of the Internet and the World Wide Web is being a scalable without limit. The scalability of both will be tested -- soon."
Pervasiveness was also evident in the final theme of the conference: web, society and health.
Mike Bainbridge is the clinical architect for the NHS Connecting for Health programme. While the technical problems of moving every UK patients health records online is considerable – possibly a bigger concern is societies acceptance of such a move. Mike sees the problem of access control to these records as two ends of a see-saw.
INSERT: 15 Mike Bainbridge
"You can have absolute confidentiality: only your doctor, or a specific person who you allow specific access to this bit, or this bit, or this bit of sees it. And at the other end you've got anyone who needs to, in the service, can see it. But one of those is unsafe, because you may not divulge the right information at the right time, because with the best will in the world, you don't have the medical training to know which bits are important. And at the other end, we’ve got the possibility of inadvertent divulging of information. So, somewhere in the middle, for everyone, is a level which is comfortable for them. And 80 odd percent of people believe it's right towards one end. There is nothing in their medical records that is, someone who is appropriately looking at their record should not see. And by appropriate I mean caring for them in some way. There are a significant number of people who are uncomfortable with that. So, the solutions that we come up with have to reflect that. And, it is a vital part that we can't enforce a one size fits all."
Mike Bainbridge ending this initial zip through the 15th international World Wide Web conference in Edinburgh.
Three more podcasts to go. And in the next one - as I said - we'll hear more from the conference about the Semantic web, grid computing and the impact of all of that on scientific research.
And if you'd like to subscribe to those - then go to - www2006.org - and look for the link to the podcasts in the main navigation.
This programme was produced for International World Wide Web Conference Committee by Bright Indigo.
To take a look at the programme notes - then check out the technology section of the Bright Indigo.com website.
But, until the next podcast – from me Peter Croasdale -
Sting
Goodbye.
SIGNATURE TUNE
Other resources related to this Podcast.
