Aug 19

How Will the Semantic Web “Think”?

Bill Erickson sent me Kevin Kelly’s TED talk “Predicting the Next 5,000 Days of the Web”. Kelly discusses the semantic web, which uses relational data to make associations between site to site, profile to profile. The idea of a converging, thinking web is a fascinating concept. Kelly calls this thinking web “more reliable than its parts”. He goes so far as to calls the thinking web “The One”.

I am not afraid of the semantic web. What I am afraid of is a portrayal it as merely unifying. The web converges but it also destroys and stratifies, and NO ONE can truly grasp how all of these points within the web fully relate. It asks us to separate the truth from our own perspective of that truth. If this Truth exists, we wouldn’t be able to get passed ourselves to see it.

So how will the semantic web think outside a bias? We can try algorithms, but inevitably money and time seem to get factored in. A company called Google seems to always put Knol pages ahead of Squidoo pages on search results, even though they do the same thing. Knol is owned by Google whereas Squidoo is not though. The web is also biased because most people in the world are not on it, and therefore could not offer their own even perspective if they tried.

Food for thought: is the web “more reliable than its parts” if 5,000 people report a story incorrectly and the one person who was actually there is totally computer illiterate?

How will the semantic web distinguish what is popular versus what is right? Can it? How can we take steps so that truth goes beyond hype?

  • http://www.silverspider.com Alex S. Jones

    I think it’s good to take another step back when discussing the future of the Web, whether it is semantic or another variant and keep in mind that humans are involved at every step. We create, organize and consume. The Web is merely transportation.

    We as a species tend to adopt that which is popular over what is right or useful, when the separation is significant (successful, charismatic dictator over those they oppress, gossip shows over documentaries, etc.) This shifts over time when the right option gains popularity or is promoted by a strong enough outside force.

    I think the value of the next significant incarnation of the Web will be in how it evens the playing field by reducing the time that hype dominates truth.

    Oh, and then there’s an entire discussion on “what is truth”, when we’re dealing with perspectives communicated via the Web – the Russian-Georgian conflict being a great example of two sets of truth that are not very easy to untangle, even for a simple time line.

    A relational Web is a powerful tool, but I don’t think it will be much more perfect than an encyclopedia – the relational mappings are connecting human communications and knowledge, so there are inherent flaws in any but the most black and white of subject matters.

    A bit of a rambling response on my part, but I hope it makes sense. I guess I would sum it up with a modification to one of your sentences above: “[humanity] converges but it also destroys and stratifies, and NO ONE can truly grasp how all of these points within [our shared understanding] fully relate.” If we can reduce the differences and tighten up the connections we will make progress.

    Alex S. Joness last blog post..Recent Links: August 10 to August 17

  • http://michellesblog.net Michelle

    @Alex

    Very good points regarding the Georgia/Russia conflict. Even w/ as much info as we have, it’s hard to determine where to stand on everything.

    Re: Reducing differences and tightening connections. This in the ultimate sense has to be the goal, but seeking truth, even if it is painfully different, is a part of this process.

    I guess I just get sick of “Apple web 2.0” design, the same buzz words, and this utopian idea that somehow the web is going to save the world. We often use the web to create. People like Al-Queda choose to use it to destroy. Do we chose to see this? How does this affect our perceptions of the world?

    I don’t want to be a terrorist, but I’m also a bit bored of it all in the end.

  • http://www.silverspider.com Alex S. Jones

    The Web, and the Net as a whole are a technology. It’s no different than paper. A quick review of history will show you many instances where writing and printing have caused mass consternation and fear as well as many instances where it has served as a light in a dark life. Technology isn’t good or bad, it’s the way that it is used and the way that it is consumed that matters.

    The semantic Web holds promise of making it easier to learn and connected topics. It also holds the promise that these connections will require little to no human effort, but a scientist searching for an antidote will be searching the same wealth of knowledge as the student writing a report on the plague and a terrorist hoping to destroy others in their search for media relevance.

    The technology in and of itself doesn’t matter. The medium doesn’t matter beyond the amount of people it reaches.

    The message matters. Whether or not it is verifiable via current means or future means, it’s the message. People will tell the truth or they will lie and distort as their goals require.

    The “thinking” Web is made up of humans with better interconnections than we currently have, but it will be susceptible to popularity, we just need to find ways to even that balance without making the system easy to scam.

    Alex S. Joness last blog post..Recent Links: August 10 to August 17

  • http://michellesblog.net Michelle

    Alex, props. Very well put.

  • http://www.pjbrunet.com/ PJ Brunet

    I don’t think finding great information is hard. How many search engines, robots, directories, libraries, organization/classification systems do you need?

    As for “intelligent agents” and “thinking”, the web does that all day long. Probably a third of all the traffic to this blog is non-human already, look at your logs, some of those bots will identify themselves as such. In fact, I wouldn’t be here right now if it wasn’t for a Google Gadget agent running on my desktop.

    PJ Brunets last blog post..Criminally Accessing MySpace

  • http://www.nmk.co.uk tim hoang

    it’s an interesting debate and what it ultimately boils down to is whether the Web can think for itself when humans struggle with many of the issues. I think the Semantic web will not be this all encompassing entity – more like a gradual step by step that makes searches (or whatever) more relevant as time goes on. It’ll be evolution not revolution

  • http://weboz.iam.ba keo

    Great post m8, looking forward to more posts like this, so i bookmarked you 😉

    keos last blog post..Link feed – submit your blog to get better page rank !

  • http://danny.ayers.name Danny

    “How will the semantic web distinguish what is popular versus what is right? ”

    In itself, in general it can’t. But what it can do is help people make such decisions. Right now it’s possible to use Semantic Web technologies to enable provenance tracking and allow statements regarding trust.

    For example, the DIG blogs (http://dig.csail.mit.edu/) allow comments only from people the bloggers have asserted they know (through the bloggers FOAF files).

    See also Tim Berners-Lee’s “Oh Yeah?” button suggestion from a few years ago: http://www.w3.org/DesignIssues/UI.html#OhYeah

  • http://www.digital-constructions.com/blog/blog.html Ed Richardson

    Its a debate that is only further emphasised by the number of comments that you’ve received at the bottom of your post.

    The web, as with many other forms of media, is often driven and then in turn delivered, in response to mass demand.

    Web 3.0 should improve the ability to find related information, but as with the current state of the web, further research and investigation should be taken before following a specific train of thought.

    Again, going back to the mass demand statement, in turn the search algorithms will only further emphasis any false “rumour” due to the nature of us as humans and our interest in “popular rumour”.

    So for those of us, that know how to pursue an information search and confirm the results Web 3.0 should be a very effective leap forward.

    But for those following gossip, it will only lead to further gossip and hearsay, but then what do those people want if not more gossip? People buy tabloids because that’s the content/copy they want to read, the web is no different.

    It is what we make of it, well not specifically “we” but you catch my drift. Corporation such as Google will always spin their take on a matter as much as a newspaper might hold more weight with specific B2B relationships and political allegiances on its editorial stance.

    Ed Richardsons last blog post..Hunt For Smart Phone Hots Up

  • http://www.postlinearity.com gregorylent

    just a cooler filing system that relates meaning(s) … my bet is it will be in chinese, which has far more meaning per character than english, is already the second largest web language, with a smarter economy, profitable and more creative web businesses, and all your children will be studying chinese from the first grade so that they can play with the big boys …

    but as to raising character, creating wisdom, no, none of that will happen .. it will be the same banality and inanity and ignorance as before

  • http://www.pabarton.com Paul Barton

    Is the web the last place of free speech or is that even being hampered. The only trouble is free speech is indiscriminate and can cause untold harm if what is said can be misunderstood – and guess what everything can and will be misunderstood. Some will even read this and think i’m for regulation – which i’m not. We need to teach people to think and reason for themselves rather than believe everything they read or hear is true. My over riding question is this “this is what i’ve learnt so far – what more can i learn in the next 3 months/3 weeks/3 days/3 hours?”
    Keep on learning and keep on asking why and the truth will come out. For example the cure for cancer is already known but not enough people are asking the right questions.

  • http://maheshcr.com/blog Mahesh CR

    Truth is subjective. The semantic web will only augment our ability to discern truth from the mass of information.

    Practically speaking, within its circumscribed limits of influence the semantic web will allow insights to be gleaned and automatic actions to be performed on behalf of its human sponsors.

    To depend on machines, the semantic web or the One or whatever, to infer truth will be a non-starter.

    What the machine can do is to do a pattern analysis, coupled with the idea of semantic understanding, and arrive at an approximation of the right position, if it can be formulated as a proposition with surrounding facts. Like a spam filter weeding out crap, this ‘untruth’ filter would parse out assertions that disprove the initial proposition and associated facts.

    That said, notions of ‘rightness’ are so convoluted and tangled within the self-interests of humans that any attempt to disembody it will be within the realm of impossibility.

    Mahesh CRs last blog post..Links for 2008-12-06 [del.icio.us]

  • http://oxfordseo.com Igor Goldkind

    I found this comment and blog because I use Google Alerts to inform me of new mentions of reference to the semantic web. Google Alerts aren’t ‘thinking for me’ nor do their notices provide me any guarantee as to the integrity or reliability of the information they direct me towards. However, they do get and receive my attention.

    And that, at the end of the day, is what the currency of the web is about: eyeballs on screens; data that summons our attentions, both individually and collectively.

    One could argue that Google’s entire business model is built on a foundation of the numbers of eyeballs on their data (your screen, your browser; their data).

    The idea of reliability or ‘Truth’ introduces a philosophical tangent to the web in general and the semantic web specifically. If I may I be so bold as to throw my hat into the discussion, I would assert that it’s ‘relevance’ not ‘reliability’ that is the motor driving the semantic web.

    Tim Berner-Lee strew the seeds of semantic development nearly six years ago as a call to answer a problem: that of organising the (even then), rapidly unfolding mass of data being published on the web into more useful and accessible data structures that would still maintain the strong democratic ethos that is the keystone of the www.

    The answer to that call to action came from development teams from Stanford to Cambridge (as well as Manchester, Liverpool and Oxford) who through submissions through the w3c reached a loose consensus on standards for meta formats (Resource Description Frameworks); as well as a structure for ontologies (OWL) . What is referred to as the semantic web is not a killer application nor a replacement mark up language but a consensus on a approach to making the existing data on the web more searchable, more accessible, more ‘useful’ by being more relevant. ‘Relevant’ in terms of the inquiries, the questions that we pose.

    Michelle’s point about Google placing Knol pages ahead of Squidoo pages on search results, as Knol is owned by Google is precisely what semantic development is poised to take on. Currently, Google controls the hierarchy of relevance for data on the web. Not because they are evil and power mad but because so may of us use them so often. And we use Google because Google is ‘useful’ and serves our interests.

    It that sense there is no ‘thinking outside of bias’ because we are biased by the themes of our interests that determine what we pay attention to. We can be however, swayed by the data that at least appears to be more relevant to our interests and in that sense Semantic is more useful because it more finely discriminates the relevance of data published on the web.
    The Semantic Web will only add an extra dimension of accessibility to the data that is already being published, it won’t be able to determine reliability, ‘Truth’ or political bias any more than Google does now.

    But it should make our answers more relevant to our questions.

  • http://oxfordseo.com Igor Goldkind

    (sorry, the link is unintended, it should read ” . . . is the keystone of the World Wide Web”)

  • Zoujiaofang