Extreme Data

Longform

Image: Stcokfresh

Even grasping what discoveries might mean is where the foray of analytics into big data is heavily dominated at the moment by data discovery and visualisation, according to Towey. “We are a long way from industrialising any of this effort with standard mathematical models. We are still exploring the topography of data in those pools and oceans. Our efforts are in many respects aimed at identifying the little pieces of signal in all the noise and then taking them into a structure for the analytics exercise.

“But one of the limiting factors is definitely the Talent, what we are now calling data scientists,” Towey believes, “That is why we are a long way from industrialisation. Look at analytics outside of Big Data and the very recent years when we have been trying to push some tools and capabilities out to ordinary business users and democratise data analysis to some modest degree. We are barely started on this journey.”

Out with the Vs
Oracle’s EMEA director of Big Data Michael Connaughton is another expert who is not keen on those infamous Four/Five Vs “I have a hang-up with people using those V things. Most of the meaningful comments and definitions I’ve heard about Big Data have been in the context of an organisation being able to actually use data. And in fact I think that’s even more helpful when we start to think and talk about extreme data. In fact ‘extreme’ in many ways depends on the context of the resources you have available to handle it. We have telco clients where an extra 100TBs of data does not break a sweat while there are large organisations where that would break the information management architecture.”

That kind of capability is going to be offered more and more by third parties, Connaughton believes. “We already have partner clients which offer ‘Insight as a service’, as you might call it, a Big Data managed services capability. As has often happened before in ICT, there are organisations that do not want to invest in the specialised infrastructure or systems themselves but they do value or even need what it can deliver.”

Read More: analytics Big Data Leslie Faughnan longform