By Péter Vaskó
in4 Ltd, CEO

Hungary is a very small and landlocked country, its people speak one of the most difficult languages, therefore Hungarians must come up with something in which their compatriots find amusement and delight. It is very likely that famous Hungarians like János Neumann, Pál Erdős, or Charles Simonyi were also motivated by the very same thing. They could, for instance, come up with an application we all miss from the World Wide Web.

The state of things

The Internet was conceived by a genius spur, but was already old when born. When browsing the Internet we may often feel that it is a far cry from the media revolution we think it is – it is much more like a spiraling arm of the Gutenberg Galaxy. The generally used means of the organization and search of information are closer to what philologists of the Alexandrian Library were doing two thousand years ago. On the shelves accessible to the average user one can find mostly closed, linearly readable linguistic units, what we may call, in fact, static books. The librarian, with colorful letters sewn on his overcoat, would dump a good pile of these in front of us, if we ask him nicely.

Wiki, the important media-innovating technology is one step ahead, with the great merit of collecting community knowledge into one location which would be otherwise dispersed in space and time. However, Wikipedia is still the offspring of the 18th century lexicographers. Information is stored in a way readable by humans, but not renewable and not transportable on a system level, since it is buried in language and cannot be exported easily. On the other hand, the “new” Knol is content-wise one step backwards: it is not collaborative, but competitive, and it encourages not cooperation, but rather con-operation. Nothing is more boring than turning back.

The G-spot of the Internet is not here. The founders of the web, e. g. Vannevar Bush (Memex) or Ted Nelson (Xanadu project) had way more daring and radical visions of what the new home of intelligence should be like, where the elements of knowledge are integrated into one another in free and varied network systems. However, it seems that all this had vanished in the mists of time.

Ted Nelson described the possibilities in data relations as the Literary Machine, and the Web is indeed more of a poetic space than that of engineering. The relations are out there on the Web, awaiting for being explored by the machine. After the force policy of Microsoft, followed by the sheer pragmatic and minimalist era of Google, the symbolist period of the World Wide Web will arrive, exploring and establishing hidden relations: a period with more surprises, more passion and excitement.

Ghost in the Shell

We experience and re-live the world as a whole, as the simultaneous and complex system of relations. This holistic experience will disappear when we analyze the individual details. This is especially true with the Web, where the search engine will rake together a huge pile of information-crumbs in the list of findings, but these will not form an organic whole.

If there exists anything like a truly exciting challenge, then it is no other than creation, using the information-mound chopped up into data, texts bits and pictures. The unearthing of those relations which connect together the various pieces of the world. The Web can be way more than an information-dump, or a pile of data, for which we are using it today.

How can we regain creativity from the success of routine search, embodied by Google? The greatest problem the challengers of Google face – and probably also the cause of their failure – is that they fell under the spell of routine search, and can not anymore think outside of the box. They are throwing in the most complicated technologies ever in order to produce a simple door-knob. The “new” search engines, endowed with tons of innovative solutions, are able to come up with the same type of catalogue-lists of the same type of closed, sealed books as the market-leading search engine. As if one was aiming at defeating Goliath in arm-wrestling.

Fighting Google is a mistake, the big G does its job well, there is no point in envying it for this. The real war should be waged for a method which can transform the quantitative increase of the data available on the Internet into a qualitative change. We are past that phase when the large majority of people can read and write via the Internet, the same way as they do with books. The real war will be waged for the regaining of creative thought and the liberation of fantasy, so that we will be able to inject the Internet with real structure and tie up the loose and chaotic threads into a true network to work with.

Correspondences

How can we step beyond the magic circle of the success of routine search rooted in experience gained through the bookish world?

While contemplating over these questions with my colleagues, we gathered a few old and new basic principles by which the world-wide web can be made a more interesting space:

- information must be made freely transferable and recyclable
- it should not be us to look for the information, but information should find us
- the machine must adapt to us, not us to the machine
- the possibility must be created for the network to be able to teach itself
- more fun: the relation of data to one another and to us must be made more creative and more intuitive

For attaining all these, we decided on creating the following tools:

- flexible and language-independent information structure
- a collaborative database able to handle semantic relations utilizing community intelligence
- an algorithm extracting relevant data from natural language texts
- context-sensitive interface capable of showing data-relations in several dimensions

The center of our experiment is a semantic database model which, besides the collaborative data-gathering, will soon be able to teach itself with the help of the pieces of information collected from the Web. In the first phase we are aiming at organizing common knowledge nodes, such as people, institutions, geographical places, concepts into a coherent system. We want to link up in a common thread the classical bottom-up approach with a top-down logic and the semantic database technology. We gained a multitude of experience from various applications and technologies (Microformats, Opencyc, Dapper, Open Calais, Spock, Freebase, Blueorganizer, Lingospot etc).

The initial result of the iGlue project experiment is the demo version demonstrating this data-flow linked into this common blood circuit. Pieces of information cropping up in the different types of media as fragments or as separate items, but belonging to the same entity, can be treated separately, as we are doing in most cases nowadays. It seems evident though that we glue the elements belonging together with an intelligent super-glue (hence the name ‘iGlue’). Redundant and fragmented pieces of information scattered around in space, time and in different types of media will thus be able to become parts of a larger whole. One piece of data or one page can not only be the end of a search, but also its starting point. This way a recognized and meaningful character-string will be able to connect to that larger network of meanings which we could call the organic intelligence of the web itself.

There may come a time on the jungle of the WWW when the new type of sensitivity will take over, as Baudelaire is also writing in his famous poem on relations:

‘As long-drawn echoes mingle and transfuse
Till in a deep, dark unison they swoon,
Vast as the night or as the vault of noon -
So are commingled perfumes, sounds, and hues.’

(poem translated by: Roy Campbell in: Poems of Baudelaire (New York: Pantheon Books, 1952)

This entry was posted on Friday, August 8th, 2008 at 11:30 am and is filed under CEO Views, Guest Authors. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.