Big Data and Ambient Analytics
The term “big data” implies that size is the single distinguishing characteristic of the raw digital chaff from which businesses must now extract the wheat of actionable information. But as cloud computing becomes more pervasive, it’s not just the size of data that challenges organizations, but its fluidity.
As organizations are making increasing use of SaaS, PaaS and IaaS, they can no longer be sure that the data they need is sitting on a known set of servers in the datacenter. In fact, because LOB users are now subscribing to SaaS solutions without asking anyone’s permission, business data can quickly accumulate outside the enterprise before IT even knows it’s happening.
Also, some of the richest big data isn’t generated by the business at all. Instead, it belongs to social media, research companies and other third parties. So IT must become increasingly adept at working with (discovering, navigating and visualizing) data sets regardless of where they are—or aren’t.
Finding the good stuff—fast
Big data success does not necessarily come from aggregating massive amounts of digital chaff from a large number of disparate sources. On the contrary, it can sometimes be completely unproductive to waste a lot of storage and compute power on processing too much data. The key instead is to discover and navigate through the useful data hidden inside big data sources—and then perform the appropriate analytics on that data alone.
The problem, of course, is that data sets of “big data” do not always offer the transparency we’d like them to have. Its metadata does not conform to our data management dictionaries. And we’re not always entirely sure what we’re looking for, anyway. IT therefore needs more intelligent tools for quick discovery, navigation and then iterative extraction of data subsets from large data sources based on both “hard” and “soft” criteria.
Empowering the business user
As pressures grow on the business to respond more quickly to emerging trends and problems, IT can’t allow itself to be a bottleneck in the big data analytics process. Sure, IT will always have a role in facilitating analytics and performing more sophisticated analytical tasks. But, wherever possible, IT should enable business users to be self-sufficient when it comes to pinpointing actionable information buried in large sets of structured and unstructured data.
Also, business users are constantly changing their minds as market conditions shift and competitors make unexpected moves. When this happens, business users need to be able to act quickly—and on their own.
Simply put, the issue today isn’t just the emergence of data that is “big.” It’s also the need for analytics that can be “ambient.” We have to empower more people to be able to more quickly point more intuitive analytic tools at data that can be more places and in more different forms.
Vivisimo clearly has an important role to play in this new world of big data and ambient analytics. And, as part of IBM, we see ourselves fitting nicely into the larger portfolio of big data analytics solutions. But it’s important to make it clear that we are not just a hammer looking for a nail. The relationship between business and information is evolving rapidly and in some very significant ways. As that information becomes more ambient, our navigation and analytic capabilities have to become ambient as well.