Sunday, June 10, 2007

Self Organizing Content

I started looking into details of the Semantic Web about a year ago for a startup idea I was working on with a couple of friends. The concept on SW and the potential of it fascinates me. It is the web where data is stored in ways that machines can understand - not just humans. From Wikipedia:

Humans are capable of using the Web to carry out tasks such as finding the Finnish word for "car", to reserve a library book, or to search for the cheapest DVD and buy it. However, a computer cannot accomplish the same tasks without human direction because web pages are designed to be read by people, not machines. The semantic web is a vision of information that is understandable by computers, so that they can perform more of the tedium involved in finding, sharing and combining information on the web.

From a business perspective this offers massive potential for new waves of development and data analysis in new creative ways. (A good article by Business week on the topic) The interesting thing is that content on the web is already starting to self organize itself through tagging. Tagging doesn't just describe 'what' but many other aspects of the tagged object. This includes human associations that are key to understanding.

For example a take a picture of a cup on flickr which may be tagged with coffee, tea, both etc depending on different perspectives. The association between coffee and cup would not be obvious (especially to a machine!), but you can start to see with a world wide community of users tagging and creating associations, it starts to form the basis for content self-organization and association. This alone, does not create the semantic web, but begins to form the kernels in which to create one.

Tom Haskins refers to this concept in a different light, calling it reflection:

Those of us who search for, link to and tag digital content are practicing reflection...We are externalizing meaning. We are connecting our dots for others to consider how to connect their dots...We call it like we see it and that differs from everyone else. There are many right answers. There are thousands of ways to see it, frame it and use it. We have final say over our idiosyncratic consideration.

The next step beyond creating associations is to begin unearthing unknown or unlikely associations between concepts both inside and outside the walls of the organization. With the startup I was working on last year, part of what we were trying to do was have software help drive innovation, by 'understanding' disparate ideas created by people. By having software understand an idea, and then begin to draw associations between disparate ideas to create new streams of innovation.

Exciting times ahead, and we are all helping the web organize itself and realize the next evolution.



It a great topic and I'm glad your bring it up. There are many fronts to giving context to content found on the web using machines only.

Microsoft PhotoSynth is one such project that can index all the images in Flickr and produce a very interesting prespective of their target tag such as the "Notre Dame".

Watch more of it here and here

Kevin Donaldson said...

Thanks for the comment and the links! Hope all is well.