Sematic web and information processing

Qwiki looks like a very interesting platform. It’s like Wikipedia in that it is like an encyclopedia of general knowledge, only instead of the knowledge being constructed primarily by contributors, it is created by machines, pulling all these little bits and pieces of content from other spots on the web. It does this on the fly using semantic web technologies. There is a way that users can participate, by suggesting sources of information that might improve a Qwiki, but the heavy lifting is primarily done by machines. And it looks very pretty. The UI is slick.

In taking a look at Qwiki, I came across this blog post from Gregory Roekens in which he connects semantic web technologies with a theory of knowledge creation and information processing called mental space theory, which, in turn, is based on something called a DIKW (data, information, knowledge and wisdom) hierarchy. DIKW illustrates a hierarchical relationship in that data and information lead to knowledge, which leads to wisdom. I haven’t come across this term or theory before, but it is intriguing.

Amplify’d from

Qwiki is one of those emerging platform leveraging the semantic web. I often used the Ackoff’s allocation of mental space theory to explain the importance of Semantic Web and its huge potential. This theory is based on the DIKW hierarchy.

In a nutshell and using the diagrams below, our brain is using 40% of mental space to process data into information, a further 30% to process information into knowledge, 20% to process knowledge into Wisdom and only the remaining 10% is used to process Wisdom into Vision (see diagram 1).

In his work Scott Carpenter explains that thanks to data-handling technology (think excel spreadsheet, charts and dashboard) it allows the human cognitive energy to shift upward and produce information out of data (see diagram 2). Without these technologies the cognitive is locked down by mundane and time consuming effort to process the data into information.

What’s really exciting in Scott’s theory is that with the Semantic Web and its semantic processing power cognitive allocation can shift to Wisdom and Vision with the machine effectively delivering the Knowledge (see diagram 3).


Qwiki via Stephen Downes


Adaptive learning, disputes, and breaking out of echo chambers

I have just installed a FireFox addon called Dispute Finder. Dispute Finder is an addon developed by Intel Research and UC Berkley that highlights disputed information on a web page and displays alternatives to that disputed claim. It uses both crowdsourcing and curated resources to try to expose you to alternative views about what you are reading.

As my Masters has progressed, I find myself becoming increasingly interested in adaptive learning systems and the role that technologies could play in shaping a users personal learning environment. Now, I am no computer scientist and when I hear words like ontologies being thrown around I have to admit my head begins to ache slightly. The depth of my knowledge of semantic web technologies doesn’t go far beyond a high level flyby of FoaF and RDF . Nonetheless, I remain interested in advancements in recommendation systems, both technical (semantic) and human (folksonomies) and the implications they could have for learning and constructing knowledge.

More and more on the web we are seeing personalized recommendations pop up for us to explore, often based on our past behaviours or, increasingly, recommendations provided to us by our social networks. Amazon recommends books to me not only based on what I have bought or browsed before, but also what other people who have bought or browsed similar titles to me have found interesting. Facebook will recommend friends to me based on who is already in my network, and adjust the information I see about that network based on my viewing habits (and some other variables, I am sure).  When Facebook introduced a real time stream a few versions ago, it did so with a News view and a Live view. At the time I wasn’t sure what the differences were, but after using it for awhile the advantage of the News feed becomes clear. The News feed is content that the system deems to be more relevant to me – it is a filter to help control the tidal wave of network information (I have Clay Shirky in my head saying “it’s not information overload – it’s filter failure“). And most of the time, it is right.

I am intrigued by what it means for learning if some of the construction of these connections is being done by technology, and how educators can assist learners in setting up environments that are conducive to this kind of semi-organic discovery. On one hand, these types of recommendations help to bring order to the chaos and may open up paths for exploration that may not always be obvious. On the other hand, they also set up the possibility of developing echo chambers. If the only information I am being exposed to is information congruent with my own views, then how can I be expected to become a critical thinker? After all, being critical often means being able to discern between two opposing points of view. How can you do this if you are only being presented one point of view?

Which brings me back to Dispute Finder and why I find this project interesting. Dispute Finder seems to depart from the general trend of recommendation engines on the web. Instead of recommending things it thinks I will like, it shows me information that may not be aligned with my own views, which opens up a possibility for me to learn.

via interview with Rob Ennals on Spark