How do you know if your big data is actually big data?
The question may seem trivial, but it is actually very important, as certain types of data are used to solved different kinds of problems. This means that a business that is leveraging data must do so in the right way if it wants usable results.
In a recent interview with CXOTalk, Dun and Bradstreet Chief Data Scientist Anthony Scriffignano, explained the difference between big data and smart data, and how businesses should go about solving their problems with data.
SEE: Job description: Data scientist (Tech Pro Research)
Big data often deals with what are known as the five Vs: volume, velocity, veracity, variety, and value. And, when any one of those Vs of your data begin to overwhelm your attempts to control it, you have a big data problem.
Smart data, however, is a little more nuanced.
"The smart data is the subset of that data that will actually apply to your problem—that can be used intelligently in a way that takes you towards a solution," Scriffignano said.Although, he said, it doesn't have to take you directly toward a solution in his opinion. It could also help you, or your organization, either break a large problem down into a smaller problem by eliminating variables, or it could also help surface a question that needs to be asked of your problem.
So, how does one begin their big data journey? For starters, you always begin with the problem that your organization is trying to solve.
"You never lead with the data, and you never lead with the technology," Scriffignano said. "You lead with the problem."
SEE: How to make yourself a DIY data scientist (TechRepublic)
Leading with the problem keeps you focused in your big data strategy. Using big data or analytical tools for the sake of their novelty has the potential to waste time and deliver bad returns on its investment.
There's also the potential for bad science. By looking for data to support a belief, or building up a strategy around proving something that you want to be true, you're not leaving room for the scientific method, or for properly testing a hypothesis. You shouldn't leave out the science of data science.
However, there is still the question of what problems you should be looking to solve with data. Scriffignano said that problems "to look at" and "to solve" are two different questions but there are some guiding principles.
Basically, the problems exist in one of two categories: Total opportunity and total risk. At Dun and Bradstreet, which analyzes businesses as potential partners for other businesses, risk could be whether a company would stay in business or threaten a business objective of their client. Opportunity problems could be how big the target company is or how much they might complement an existing customer.
When it comes down to it, don't look to simply solve cool, or unique, problems—focus on problems that affect the bottom line and problems that add value to the business.
Also see
Source: http://techrepublic.com.feedsportal.com/c/35463/f/670841/s/4ec9cc5e/sc/23/l/0L0Stechrepublic0N0Carticle0Cbig0Edata0Evs0Esmart0Edata0Edun0Ebradstreet0Echief0Edata0Escientist0Ebreaks0Eit0Edown0C0Tftag0FRSS56d97e7/story01.htm