The Googlemonster’s unveiling of its new ‘Knowledge Graphs’ makes perfect strategic sense for them and their finances. And it also makes huge sense for anyone trying to conduct a sensible search these days and get real value.
Since search became the dominant knowledge acquisition paradigm in the late 1990’s, there have been attempts to find some ‘meta’ search method. This usually involved parallel search and subsequent disambiguation – anyone remember Mamma?
But the huge power of Google and the simplicity of presenting results ranked by relevance to the search query largely made those attempts pointless – until recently.
The WWW set information free and made possible the inexorable process by which all human knowledge moves online and is created online – a huge attractive force like a reverse black hole. The volume of information is already overwhelming and growing exponentially. It’s all too much. So today, searching for anything is like trawling the top centimetre of the ocean for fish – you’ll get a meal but did you really find the best you could get? You won’t know unless you read x millions of individual results.
So Google attempted to create context by using algorithms to refine and prioritise the presentation of results based on prior user behaviour. But that has obvious conceptual limitations: finding the right type of information for me NOW is not necessarily optimised by what I searched for before, especially if I am searching for something entirely new.
The new knowledge graphs approach the problem from the other end by effectively offering a form of visual multiple choice. By doing so they bring back again the serendipitous joy of search , finding information you never knew you wanted – and connexions which you never knew were there.
They also re-introduce the whole idea of visual search – an idea which Microsoft’s Bing and newcomer* FidEq have also looked at as a more natural way for visually programmed humans to filter large amounts of information in pursuit of understanding. Before language there was sight.
Infographics will never be the same again – I hope.
* Disclosure: Tony Burgess-Webb is a minority shareholder in FidEq, which owns a US patent for visualisation of network connections.