Friday, February 22, 2013

The Army's need for Data Mining


The Army’s need for Data Mining



                During this critical time in US history, with the financial crisis, the Debt Ceiling and potential Government Sequestration, the theme around the Pentagon and around the nation is “Do More With Less”.  With budget cuts looming and the obvious downsizing of the DoD, the task at hand for our military leaders is to determine how to maximize our investments. Obviously with less money to spend, fewer men and women to perform the task, minimizing inefficiencies is the key. Just within the Army, they have identified many opportunities for savings in application modernization. Within the IT operations realm, the Army can eliminate many program and data duplication. Col. Chris Miller (the Army’s data center consolidation program director) believes that they can remove this duplication and prepare applications for movement into other operations environments. With the focus of two wars for ten plus years, there hasn't been an emphasis on long term planning. The priority was to supply the war-fighter with what they needed as soon as possible at the cost of managing and maintaining these quick fix databases. The Army has found at least 16,000 applications throughout their station’s networks and post camps that need to be modernized and synced together.
                These data sets need to be mined for common file types and names, duplicate files, lessons learned from how the data got duplicated, and other useful information. The lessons learned can be significant for future use of the networks and servers. The effective planning of how data is stored on these servers is very important, because the Army is planning a reduction from 500 data centers down to 185. The Army is still doing a discovery and inventory of all data centers. Miller stated “It is important for us to get the figure right on the number of data centers we have so we can attack, virtualize, close and modernize our inventory to get it into an environment that is manageable”. The Army realizes that there are technical solutions to aid with their duplicate data problems and consolidation efforts as they move to cloud computing. Data Mining tools and techniques from this class could be used to help consolidate these data centers. Miller stated that “Technology, we know it is out there. That is the easy part. The bigger challenge is the culture and politics.”

                GCN reports that the Army may not have the “BIG” data sets of some commercial IT organizations, but the data it does have is complex. These complex data sets have useful and viable information linked in, but it requires a set of skills and techniques to mine this data to retrieve its useful characteristics. After this enriched data is found, the Data Analyst would be tasked to enrich and expand it for quick real time analysis for the decision maker’s specific missions. This data could be placed in an easy modified visualization for inference into various scenarios. There is a saying of “The answer is right in front of us” but the decision makers need someone to find it for them.  

1 comment:

  1. Jason,

    Very interesting. Additional references:
    http://www.wired.co.uk/news/archive/2013-01/21/equivital-black-ghost
    http://www.forbes.com/sites/ciocentral/2012/12/26/big-data-in-the-enterprise-a-lesson-or-two-from-big-brother/
    http://www.agc.army.mil/Newsrelease/OTFL_CMB_FEB12.pdf
    http://www.armedforcesjournal.com/2012/10/11458170/

    Fadel

    ReplyDelete