Peering into the Coming Year
Earlier this week, IBM unveiled its yearly "5 in 5" list - five predictions that forward-thinkers expect will come to be within five years. This year's list envisions computers that will literally "touch" us, as well as see, hear, taste and smell. As Popular Science put it, "Siri suddenly seems quaint by comparison."
Since the five-year view is taken, I turned attention much closer, asking several regular Big Data Hub contributors for their prognostications of what will happen in the world of big data in 2013.
New analytic methods will be discovered while wrestling Big Data to the ground, and they will get re-purposed and become useful on Small Data. Which is all most of us have anyway.
IBM Fellow and Distinguished Engineer
Chief Scientist, IBM Entity Analytics
Governance & Security
Privacy, security and “big data ethics” will dominate the conversation. These topics danced around the edges of the chatter at Information On Demand 2012, but now that businesses and governments are putting more big data projects into practice, a critical mass of people are starting to ask more pointed questions that we have not addressed directly or answered completely. Big data has tremendous potential to improve outcomes for citizens, governments and businesses, but only if the public at large feels secure and comfortable in the way it’s being used. Everyone has a stake in this discussion, from policy makers to technologists to companies and citizens.
Security from and for big data environments will be hot. “How can we leverage big data environments to gain security intelligence in real time? How can we also make sure we build security controls into all big data platforms?” Organizations require this focus for a wide variety of reasons, one of which is clients demand accountability for how data is used, another is compliance mandates, and yet another is internal governance controls.
InfoSphere Guardium & InfoSphere Optim
For a long time we were focused on perimeter network defenses, and while that was not perfect, it was sufficient for many organizations. However, the world is changing. Today, operationally and technically advanced attackers can get around the defenses of point products and, as a response, successful security organizations are starting to move away from the traditional defense-in-depth mindset and embrace an approach more focused on counter-intelligence. This change means collecting and analyzing more and more data, from more sources, all the time. We have some clients that see 2 billion security events every day, and their challenge – every organization’s challenge – is to find and correlate the data and events that matter. As we continue to expand the scope and use of things like network traffic analysis, contextual identity data, application vulnerability data, audit data, etc., the amount of data we’ll be forced to sort through is only going to increase. It will be the challenge of tomorrow’s security team to embrace not only securing data, but deriving security insights from the data.
We recently published a video on security predictions for 2013, and it addresses the Big Data + Advanced Threat story.
Communications, IBM Security Systems
In 2013, attention will shift from technologies of big data management to creating business value. By exploiting integration and business intelligence products already in place, organizations will bring new information sources to existing business processes and so do things better, and design new processes utilizing new information to do different things.
My prediction (tongue in cheek, though it is):
In 2013, the online humor publication “The Onion” will run an investigative journalistic series probing the challenges of corporations finding social media needles among haystacks of needles, and quickly come to the conclusion that there are too many needles to count and who cares what all those whiners think anyway? In all seriousness (well, mostly), companies will find themselves drowning in an ever-rising tide of social media conversations and begin to get much more serious about enterprise-wide social intelligence operations, which will in turn require a ramp-up in internal social listening skills development (and external consultancies), as well as increased RFQs for big social tools and technologies. Those organizations that make these investments early and often will begin to distance themselves from their competition (as some have already begun to do, but I won’t name names).
Granular opt-in policies are a “must” before location-based offers can take off in order to honor subscribers’ privacy. For example, subscribers need to be able to control when and how they receive real-time location-based offers (I want to receive offers when I want them – while shopping at a mall – but not when I don’t – while driving to work by the mall every day. Subscribers would want to control the time window, specific location or location types (commercial vs. residential), etc. 2013 will see retailers, service providers and manufacturers working closely with service providers (A&T, Verizon, Vodafone, T-Mobile...) to provide subscribers with granular opt-in policy options.
Big Data was The Word this year. So, most folks were just trying to figure it out, how it impacted their business. I think next year we’ll start seeing the fruits of this as folks deploy solutions and see results. In healthcare, the transformation of healthcare will be data-centric (I paraphrase Ray Campbell, Exec Dir, CEO Massachusetts Health Data Consortium). And whomever doesn’t realize that will be left in the dust. Or, to paraphrase Randy Thomas, from Premier: If you are not data-driven, the viability of your organization is in question.
With the pressure for changes coming from patients, payers, and government, hospitals MUST step up their analytics efforts to provide more cost-effective, patient-centric, evidence-based, accountable care (how many buzz words can I fit in one sentence!). And hospitals are woefully behind. But I suspect we will have many success stories over the next year.
And it’s not just the operational descriptive analytics (as in, understanding how I’m doing today). Predictive analytics in healthcare will be on the rise - being able to predict risks, outcomes and readmissions BEFORE they happen, and so informing choices of programs, treatments and patient management. This is what healthcare organizations will be looking for as they get their regular analytics (data sources, reporting, processes) in order.
The use of big data analytics will gain a great deal of momentum within government agencies, as they struggle with bringing down the cost of healthcare and government entitlements, particularly as it relates to fraud, abuse and errors. An increasing number of government organizations are realizing that combining unstructured data, such as social media and medical and social services case notes, with traditional sources of data can greatly improve fraud investigation efforts. These types of data and analyses will eventually become a standard part of fraud investigations. Some larger agencies will begin to implement near real-time fraud analytics, with the goal of preventing fraud before payments are made.