There’s a recurring theme in several recent articles about big data analytics. The power of predictive analytics applied to the vast stores of data generated by social media, geolocation, video and audio feeds have inspired some to ponder the potential misuses of big data.

Big Data Analysis and War

What happens when the power of predictive analytics is applied to human warfare was highlighted recently in a post by Irfan Khan, Strategic Weapon: Unstructured Data Delivers Battlefield Edge.

Reseachers built a computational model from the large store of unstructured data contained in some 92,000 individual documents – secret U.S. military incident and intelligence reports made public by WikiLeaks.

Using sophisticated spaciotemporal modeling techniques, analysts were able to predict with startling accuracy where conflicts would occur in Afghanistan, as well as the intensity of the violence. Of course, one’s political perspective will influence how you view this application of big data analytics: is it a strategic advantage or an ominous threat?

The Seven Deadly Sins of Data Science

On a lighter note, Josh Williams, President and Chief Science Officer of Kontagent, presented The Seven Deadly Sins of Data Science at Kontagent Konnect 2012.

His review of the most common transgressions committed by companies who analyze large volumes of data is wickedly charming.

He challenges his audience to repent of these classic data sins:

1. Sloth: Lazy Data Collection

2. Negligence: Misapplied Analysis

3. Gluttony: Too Many Reports

4. Polemy: Data Definition, Use Disagreements

5. Imprudence: Jumping To Conclusions

6. Pride: Decision-DrivenData Making

7. Torpor: Learning And Acting Slowly

Jeff Bertolucci’s Information Week review of William’s presentation is an engaging read — if nothing else, it has inspired me to work the word “Polemy” into my next business presentation. The sly humor of the seven deadly data sins conveys some fundamental truths about data analysis and human fallability.

Big Data Means Bigger Risk

Of course, the issue of data security is no laughing matter, as John Karabin, Area Vice President of Verizon Business, makes clear in his article on Big Data, Bigger Risks. He offers these basic security recommendations to save enterprises from the heartache of a serious data breach:

  • Eliminate unnecessary data. Unless there is a compelling reason to store or transmit data, destroy it. Monitor all important data that must be kept.
  • Establish essential security controls by ensuring fundamental and common sense security countermeasures are in place and that they are functioning correctly. Monitor security controls regularly.
  • Place importance on event logs. Monitor and mine event logs for suspicious activity.
  • Prioritize security strategy. Enterprises should evaluate their threat landscape and use the findings to create a unique, prioritized security strategy.

Although many organizations may balk at the potential expense of such increased security measures, a single data breach can easily exceed the cost of a good security program many times over.

So what’s the take-away from all these cautionary tales?

While big data offers big opportunity, it also presents significant risks. Analysis of big data must be handled with care or innovative new uses can easily turn to destructive abuses.