London mashup* Semantic Web or Web 3.0?
Thanks to an invitation from Vecosys, I gave a keynote presentation at last night’s mashup * event.
It was held at BT’s astounding conference centre, which possibly has the best conference facilities I’ve seen. It was fully equipped with a control room, 2 video cameras, live bloggers, investors, entrepreneurs and Semantic enthusiasts.
I don’t know who I was scared of most, bloggers hitting the enter key as you took a pause, or the 3 people wearing ties (Jag ?) 
Although I’m not sure how many logged on to blog given that BT were charging for Wi-Fi – how bizarre.
The evening was facilitated and kicked off by Sam Sethi who gave intimate demonstrations on Microformats. Mark Birbeck from X-Port was the first speaker up and gave a good demonstration on how his
next-generation semantic web browser that seamlessly combines XForms with other languages such as SVG, MathML and X3D.
I spoke about how Content Labels can (and will in my opinion) be a major player in changing how search works in the future. I talked about how they are a great use case for demonstrating the benefits of the Semantic Web in a meaningful way for end users. I also gave a demonstration of a Firefox extension which did all of the above.
Thanks to ReadWrite/Web Route79’s Flickr for both photographs. I believe the video will be on YouTube soon.
In essence, our Firefox extension (built by our friends at Glaxstar) provides users with more information about a particular Web site from the search results screen. This means users no longer have to hit a site before finding out if it is suitable to them. The suitability of content are determined by the users needs, such as suitability for minors, disabled users, privacy conscientious users, security, identity and so on.
If you want to learn more about ‘how’ Content Labels look to end users, visit our Search Thresher Web site. Alternatively you can read what ReadWrite/Web had to say as they’ve given a pretty good explanation themselves (well done!). There is more to add so I’ll write it up soon.
Tony Fish was last to articulate his thoughts. I’m not sure if I got what he was saying, but it sounded interesting. I’m a very philosophical person myself. Given that I own Michelin rated restaurants in Dublin, I found his analogy using food interesting… personally though, I don’t care what a kebab looks like after a few beers – watch out for the video if you want to put this into context 
We ended up with a lively debate during the panel discussion with questions flying in from the audience. There were a couple of people however, who just didn’t get what we’re trying to achieve with Content Labels. I’d like to put it down to speaking for such a short space of time, so it’s impossible to appeal. to everyone.
There’s only so much you can do to ensure you appeal to as many listeners as possible. I certainly wasn’t there for a debate about ‘RDF ‘. I was there to show what you can do with it for end users. Having said that, I did enjoy the banter with the few individuals that got hung up on police forces patrolling the Web and their definitions of machine-readable vs machine-understandable.
Basically, I was being blamed for trying to ‘police the Internet’ and ‘restrict access to the open Web’. Well, nothing could be further from the truth. I would simply like Web site owners to provide more information about the content on their Web sites so users can make informed decisions before entering them. We can also enable Web site owners to have their assertions about content verified. This helps them to demonstrate an additional layer of trust to consumers.
I really did enjoy the evening as I love a good debate. There’s nothing worse than attending an event and not getting the opportunity to ask the speakers direct questions. I hate attending broadcast events where speakers stand up, talk and then sit down.
Perhaps the people who disagreed with my approach to enabling trust on the Web could articulate their thoughts on this blog if they happen to stumble upon it in the future.
To get a very good break down on the topics that were being discussed take a look at ReadWrite/Web’s post . Thanks for the mention!
Take a look at Paul Miller’s post if you would like very good coverage on speakers and comments coming in from the audience.
The networking part afterwards was great for me personally. I met a few very interesting people who I hope to see again in the future.
31 Responses to “London mashup* Semantic Web or Web 3.0?”
Leave a Reply
[...] Read More Paul Walsh [...]
Well Tony certainly does have a very ‘sensory’ website, I think anyway. Is that sensory? I’m not sure.
http://www.tonyfish.com/
I’m not sure what to make of that. It would be great to get an opinion from Tony though. Good to see you again last night Robin! The tube got delayed and I ended up missing the car park by 5 minutes. So, ended up catching a train at 1:30 from Wimbledon and now I’ve got to train it back to pick the car up!
gah! I made it in the door around 12-ish, i think.
Robin, you’d got a strange looking flag for your location. Are you doing somethng intelligent to mask where you really are?
Hey Paul, sorry I missed the talk, I’m sure Sam put on a great evening.
About the whole content label and policing the web, I’ve got a one-plaster-fits-all response to that debate
The way I see it is that content labels are totally “user-centric” that is to say, forget about the technical aspects that its the content-site that has to publish the label, in reality its all about the end user that uses the labels. If that user is not you, then don’t sweat it, don’t use them , you don’t and won’t have to.
Its not about policing the web, thats the one-size-fits-all strategy, its about websites being able to convey their essence, or their abilities (large fonts, suitable for colour blind people) in a way that just gets out of the way of the end user, and ends the guessing game of search algorithms.
The semantic web I see is one on “even-more” choice, derived from even better use of technology, content-labels are one of those better-technologies, simply put, its about content being able to tell machines what it is instead of machines trying to figure out what it is. Makes sense, right, so how come the horses get spooked with this policing the web thing ? Its all sounds very 1997 to me
We all know that the concept of content labels has not worked in the past for things like child protection for instance, as in order to do so would mean the whole world would have to label their site as child friend or foe, so thats not plainly not going to work in that context. Plus child protection is about trust, and trust is like beauty that determined by the beholder.
But where I think you guys at Segala have got it right is that you look at promoting content labels via marketed forces, with search being the primary driver, enabling what Sam talks a lot about, better discovery and less search.
In short RDF-CL is simply a way for people to discover more relevant content using options and settings that’s the personal web right? So what is the “open web†anyway? Think about it, its what the search engine algorithms want to show you, that doesn’t sound very open to me.
The whole semantic web and content labels issue is so user centric, that, its impossible to actually object to the technology, because those that want it will use it, and those that don’t won’t. Simple as that.
Search thresher just jumps the waiting curve to show what search engines could do one day to make RDF-CL enabled search better for the end user.
- Ian Hayward
Work is routed through Sweeden. So was Ericsson when I worked there, but for obvious HQ reasons. Don’t ask me why here is!
Hey, my good friend Ian,
Perfectly put Ian. I think the guys who complained about the policing bit didn’t understand the purpose of Content Labels.
Adrian’s wife put in a way I thought was cool
I thought it was brilliant and I’m sure she put it in a way that made even more sense. I think I’ll get a quote and stick it on the site. Perhaps I should use a picture of a label on the back of a soup tin (seriously!).
Michelin restaurants?! Hehe – let me know which ones
As I start to explore Dublin’s culinary horizons, I’ll always prefer a recommendation (or an acquaintance’s own!)
Hey James, check out http://www.jaipur.ie – we’re going to redesign the site and base it on WordPress so please ignore the current site for now. Have you moved over for good?
Are you steering clear of our web vs wap debate on purpose?
Wish i’d been there last night when the machine readable vs machine understandable debate was raised. There isnt much to debate IMO…
RDF is data. Nothing more, nothing less. Machines can read it because it exists. In order to understand it though they need an enabler. Thats where Segala’s Search Thresher plugin steps to the plate. It provides the logic that allows a mahine to understand the data.
RDF on its own isnt going to make our lives any easier. It needs implementors to create the enablers. The search engines and browsers are the obvious candidates but they are not the only ones.
The Semantic Web Education and Outreach Working Group (SWEO) [http://www.w3.org/2001/sw/sweo/] is currently putting together a range of use case to demonstrate the flexibility and advantages that RDF can provide. So watch this space.
Pretty straight forward really isn’t it
…. Isn’t it???
[...] Original post by Paul Walsh and software by Elliott Back [...]
Paul, in a hurry so a few potentially scattered thoughts:* The picture you used at the top of your post is actually mine! As were the ones at Read/Write web> If you look at David Lenehan’s article, he indicates that the pics came from Route79′s Flickr ( http://www.flickr.com/photos/route79/ ) which indeed they did.
I’m really chuffed about that, because those pics were taken in a real hurry, with a cameraphone, in low level light, with no flash, with one hand (my other hand was holding my little notebook), and with no post-processing in Photoshop. In my view that’s a real testament to camerpahone photography. Just look at the colour and contrast! * Re wearing the tie – I was working in Slough that morning, and that day required it.
So when travelling down to BT Centre I had no opportunity to get changed. Plus no tie is so dot com.
* Agree with you that the venue was superb! Sorry couldn’t hang around t chat afterwards. I think what you guys have done re Content Labels looks really promising, but I have to say that I really do think that it will work best for sites that make promises about something for which THERE CAN BE NO AMBIGUITY – Adrian’s wife’s point hits it right on the head: there can be no ambiguity about the food content of an item at the supermarket. (I’m assuming that the Food Standard’s Agency in UK periodically verifies maufacturers claims) – but making promises about things such as “what’s non-offensive for certain groups etc” is so open to interpretation that if you take things like that to it’s logical conclusion then you will end up with a world full of “independent” authorities where you the user have to decide which belief system you align to and which you don’t. And this is no different to the “brands” phenomenon that we see today. Newspaper brands are a great example, many people have an affinity to the views of certain newspapers.
An alignment in belief and philosophy etc. This affinity is something gained and established over time. The loyalty to a brand and its values are earned and built up over time. Take “The Sun” and “The Gaurdian” – I believe that as brands they have evolved to become their own “content labels”, we know what to expect when we read them.
We like them. We hate them. In a world where promises are made on *opinions*, the brands themselves become content labels. In a *web* world full of Content Labels reflecting *opinion*, then I cannot easily see how they will be differentiated from “domain name” (or “logo”, or “brand” etc) . However, I CAN easily see how promises made about UNAMBIGUOUS things can server very useful functions that don;t exist today – e.g. I want my search results to only give me CORGI registered Gas installers, or websites that are designed for Mobile Phones, or web sites that comply with DDA and so on. Gotta go. Great event! And great talks from all the speakers – even though Tony Fish’s speech made me laugh!.
Help! I left a comment here yesterday but it seems it has not been published. Perhaps it is in the “spam” folder of your WordPress console?
I’ve found it Jag. Looks like the format might be a little messy due to the editor we tried out. I’ll publish and reformat for you
Many thanks Paul! I now remember what I said! Cheers.
Jag – really sorry about the reference to ReadWrite for the pictures – I’ll correct that. Hope you don’t mind me using them
Cameraphone? I’m seriously impressed given your distance from the stage!
Re “HERE CAN BE NO AMBIGUITY” – I agree! In fact, we talk about Content Labels for making assertions about conformance to standards and codes of conduct. Let’s use other methods for making other types of assertions. BTW, we are active participants in another W3C working group responsible for creating a machine-readable method for recording test assertions. Then you’ll be able to see how Content Label’s claims are verified, by whom, when etc.
Re “what’s non-offensive for certain groups etc” – I can now see how my comment can be seen as a sweeping statement. What I meant was that someone may come up with a code of conduct which represents assertions that are verifiable. I don’t see Content Labels being used for anything else – this should become more obvious as contentlabel.org starts to really take off. We’re just building the wiki and writing the vocabulary for Creative Commons and PEGI.
Regarding your point about affinities – I agree totally. I hope over time, Segala will be considered as the labelling authority for accessibility, mobileOK and other codes in the future. Just as VeriSign is for security and more recently, identity.
I would also like to see Segala become the labelling authority of choice when it comes to using its technology to enable other labelling authorities and organisations who wish to implement content classification. For example, O2 could use Content Labels for imode content to improve discoverability through search.
Going back to the affinities, we can demonstrate through the extension how to enable end users to rate the trustworthiness of claims. We’ll deliver new functionality through the extension with monthly builds to help demonstrate how the main players should be doing it by themselves.
I’m going to write quite a detailed post about Content Labels so people have ‘a’ definitive reference point that’s user friendly. We helped to write the official W3C stuff but it’s more techie than I care to reference here.
BTW, the o2.com domain is labelled for accessibility compliance and many of the microsites on o2.co.uk are also. One of my presentations shows, O2, Voda, Virgin and T-Mobile in search results – O2 is the only one with a green tick. So, you may not trust the brand more, but you may trust the relevance of the search result based on your individual preferences.
Relevance is where I see this going, but where better to start than with trust. Where better to start trust than with trustmarks which are, as you put it, not ambiguous.
So, in summary, Content Labels enable site owners make assertions about conformance to standards and codes of conduct
Isn’t it funny how some posts require such detailed comments that turn into posts themselves?!
I’m new to content labels, so I had a quick look at the O2 site. Clearly such labels are not currently machine-readable (unless your machine figured to launch the pop-up window that contains the certificate).
So does one not typically label the content itself, in a self-describing way but rather create another page that makes assertion about the page you just left? Why not?
Finally, how accommodating did you find the ‘host’ organisation in having another company’s brand as the link to click? I remember (mostly in the old days) sites with links to W3C validators. Maybe it was geek coolness to show one could write well-formed code… but having a for-profit company’s logo on a commercial site in the 21st century is a slightly different proposition.
Thinks average user: “Segala Certified? Is that the parent company of Telefonica?”
(Although I’m sure Paul, you don’t have a problem with people thinking that!)
James – O2.com has a trustmark, clicking on the trustmark as you probably did, invokes a visual certificate. The visual certificate is for users who want to read about the conformance claims they are making about accessibility. The certificate also provides users with an option to report a misuse of the trustmark if they disagree with the assertions made.
The Content Label isn’t visible to end users. It’s a file that resides on Segala’s server. Rather than add a link tag from o2.com pages, they made a small tweak to the server. So tools that recognise the tag will point our Segala for metadata information. The label contains the same information that the certificate has, but in metadata format.
It is possible for site owners to make self declarations using Content Labels. In this instance the label will reside on the content provider’s own server. I think nearly every code of conduct for which we would like to see a Content Label, should support self-declaration and independently verified labels. Creative Commons is one use case that doesn’t require verification.
As realised by self-labelling authorities such as ICRA, self regulation doesn’t work most of the time. Or at least, there is a requirement to provide verified labels also. If we permit self-declaration only, such conformance claims will become less quality driven as the number of incorrect claims increases. You know what it’s like within the world of testing – without testing, quality decreases for lots of legitimate reasons.
Self-labelling must be possible to ensure you get mass adoption for implementations such as mobileOK and .mobi
It’s about giving end users a choice. I would never recommend imposing verification as it would be a barrier to mass adoption.
Regarding the branding thing, I hope recognition of the Segala brand will be increased over time. We’re now seeking external investment to help here. Most end users don’t know who VeriSign is, yet, they enable trust for security on Web sites that they spend money on. Since buying GeoTrust in 2006, VeriSign practically has an monopoly for SSL certificates. This is far from ideal, but it demonstrates that making money isn’t a barrier. I’ll be looking to provide much cheaper alternatives to VeriSign’s monopolistic certificates
Going back to profit, Segala will soon offer Content Labels for stuff such as Creative Commons. These will be free of charge naturally. Not everything will make Segala money
but it does help to build the ecosystem and enable mass adoption.
Hi Paul,
Well done again for Thursday night and thanks for commenting over on my blog. I agree with the point made by Jag, that Content Labels will work best in unambiguous situations and feel that they will be very unpopular when they are representing ambiguous positions that are subject to belief systems. I felt that your presentation would have benefited further from recognition of this early on.
I also feel that the audience was perhaps more critical of your presentation than expected because of the additional apparent need for compliance. I know from personal experience that many developers are already finding the pressures of complying with all the different requirements of varying browsers and standards, frustrating. A more user-oriented audience would appreciate the benefits of Content Labels more than developers who it might create more work for!!
Keep up the great work!
Hear hear David on his point about “compliance” I think that (along with green ticks and red crosses) this type of thing has the potential to evoke emotive reactions. And it seems that it did. I appreciate that it’s hard to get a good point across but I think you did reasonably well Paul, if only because your passion for it came across clearly!
David: great to “see” you here after saying Hi after the event! Love to have an extended natter some other time too!
Forgot to say Paul: no worries about the picture credit – it traces back in the end anyway – and in any case I was just really pleased that you used them!
Yes, I was surprised at the quality myself given the circumstances in which I took them. But it does go to prove a point that cameraphone has got incredible potential to be a user’s “main camera”. With “proper” cams being reserved for “special occasion” use.
Thanks from dropping by David. I agree with your point regarding the use of Content Labels for making assertions that are not ambiguous. That’s exactly what we have in mind.
It’s difficult to appeal to everyone all of the time and it’s even more difficult to explain everything in such a short space of time. I did get some positive feedback from the likes of BT who would like to progress with a discussion. Perhaps most of the audience was from the development community though, so I’ll keep that in mind for future reference
So, I agree with everything you say. I’d love nothing more than say, 40 minutes to talk about the entire ecosystem that I’m looking to encourage – next best thing is a post I’m going to write here. Please do keep an eye on the blog and comment with your thoughts – it’s great to get feedback!!
What’s worth pointing out is that Content Labels can be used to highlight whatever it is that developers already comply with. That is, I’m not looking to impose anything. I’d rather give developers a reason to comply with standards, or at least the opportunity to have a unique selling point.
In fact, if you take a look at our partner programme, it’s made up of established agencies and freelance developers who use the trustmark (Content Label) as a unique selling point. That is, they can help their clients demonstrate their commitment to accessibility standards. Content Labels aren’t putting additional pressure on them – quite the opposite. Take a look at what they have to say http://segala.com/partner-programme/partner-member-testimonials/
Jag – don’t get hung up on green ticks and an red X. These are all things that are subject to change based on what people want to see. It’s getting a reaction which means I’m doing what I set out to do
It was a pleasure to use the photos. You’ll notice that I’ve changed the original post to include a link to you.
[...] London mashup* Semantic Web or Web 3.0? There’s an interesting debate going on about Content Labels following my keynote presentation at the London mashup* (24 comments) [...]
Well done on the talk Paul – I’d love to have been there.
Content Labels are a great idea esp. considering how easy they are to roll out. It was a piece of pi** adding them to my sites once your guys had done the certification.
Cheers,
Tom
Thanks Tom. We’ve been labelling sites after certifying them so site owners and developers don’t even know about the Content Label most of the time. Now that we’re promoting labels specifically, we’ll have to demonstrate just how easy it is. Watch this space – well, you’ll be one of the first we’ll tell
[...] A much needed ecosystem to enable better collaboration is starting to form in London, with people like Saul Klein kicking off initiatives such as OpenCoffee, Robert Loch and Paul Birch running (compelling) Internet People and informal gatherings for Creative People (supported by me from a BIMA perspective) and Sam Sethi running mashup events. And these are just a few of the networking opportunities that attract an average of 80 likeminded opportunists on a regular basis. Then you have intimate and private breakfast meetings which I’ve attended, along with people such as Paul Birch, Robert Loch, Olly Barrett, Saul Campbell, Judy Gibbons, Judith Clegg, Michael Smith, Saul Campbell, Sam Sethi, Daniel Appelquist and more. [...]
[...] A much needed ecosystem to enable better collaboration is starting to form in London, with people like Saul Klein kicking off initiatives such as OpenCoffee, Robert Loch and Paul Birch running (compelling) Internet People and informal gatherings for Creative People (supported by me from a BIMA perspective) and Sam Sethi running mashup events. And these are just a few of the networking opportunities that attract an average of 80 likeminded opportunists on a regular basis. Then you have intimate and private breakfast meetings which I’ve attended, along with people such as Paul Birch, Robert Loch, Olly Barrett, Saul Campbell, Judy Gibbons, Judith Clegg, Michael Smith, Saul Campbell, Sam Sethi, Daniel Appelquist and more. [...]
[...] A much needed ecosystem to enable better collaboration is starting to form in London, with people like Saul Klein kicking off initiatives such as OpenCoffee, Robert Loch and Paul Birch running (compelling) Internet People and informal gatherings for Creative People (supported by me from a BIMA perspective) and Sam Sethi running mashup events. And these are just a few of the networking opportunities that attract an average of 80 likeminded opportunists on a regular basis. Then you have intimate and private breakfast meetings which I’ve attended, along with people such as Paul Birch, Robert Loch, Olly Barrett, Saul Campbell, Judy Gibbons, Judith Clegg, Michael Smith, Saul Campbell, Sam Sethi, Daniel Appelquist and more. [...]
This content has been Agglom (erated) with other similar ones on http://www.agglom.com/agglom/82 – Web 3.0 – meaningless or future – What do you think about?