I had the dubious honour of attending a recent IBM-hosted event on web science, where the great and the good gathered to explain how they are finding new ways of studying the web in all its complex glory. The Web Science Research Initiative was heavily represented at the event. This organisation was set up in late 2006 by the University of Southampton and MIT to promote the idea of an academic discipline, a “web science”, blending the sociological and the technical, and to lead in research, thought leadership and education. As Southampton Uni’s Wendy Hall said, the idea is not necessarily that they have all the answers but rather that they begin to formulate new methodologies for discovering them.
There was an awful lot of theorising going on, in particular by the arch web theoriser himself Tim Berners-Lee. He praised Google for enabling people to find stuff on the web, and discussed how the web of links developed, where success is based on how many sites yours links to and which ones link to you. And then came the semantic web; this is really Tim’s baby, and what he envisaged the web to be from the very start, where “the links aren’t the most important thing but the things on the other end of them”.
The semantic web has long been heralded as the future; it’s about having a web of data rather than a web of documents – which machines can process and make sense of. Creating a web which machines can understand in this way is going to take a while, but some organisations are already using these principles internally to organise and tag content in order help information workers find relevant information more easily. And now Yahoo has announced it is to embrace some semantic technology standards there seems to be a real sea change around how firms are organising and tagging their data assets.
The knock-on effects could be massive; not just for web search either, although this may well be the killer app the semantic web needs to jump into the mainstream. It could mean web search results actually become relevant – instead of returning a long list of results, the technology will be able to recognise the object being searched for and aggregate information around it.
I really think we are just scratching the surface of what this is going to become. But in the same way its reasonable not to get to excited, as there are immense complexities to understanding text. How many people come away with the same set of recollections after reading the same passage? Hoping computers will magically created the 'truth behind the narrative' won't happen any time soon IMO.
ReplyDelete