Big Data, Predictive Models and a Hit TV Series (Part 1 of 2) by Jon Hansen

Posted on May 22, 2013


“What you are talking about is the utilization of a predictive model similar in many ways to the system featured in the hit TV show Person of Interest.  Quite simply it is a centralized capture and correlation repository which then utilizes advanced algorithms to put an organization ahead of the demand curve.”

The above is a comment I made regarding a recent USA Today article by Jonathan Golovin titled “Big Data’s lessons for old school in new media.”

Golovin talked about the real-time tracking of “spikes in sales . . . at stores in upstate New York,” as a means of turning big data into a proactive tool that would enable companies to “send hyper-targeted, online or mobile coupons” for specific products to customers in a particular region.  However I could not help but smile when he made the personal observation that “Thanks to recent advances in Big Data and expanded retailer data-sharing programs,” he did not think that “these scenarios (or capabilities) are that far off.”

The fact is that predictive models such as these are nothing new, having even been recognized in the entertainment world through the hit television series Person of Interest.

bullying person-of-interest

In the show’s storyline the main protagonist – Harold Finch – built an advanced computer system for the government as a result of 9/11.  The system uses the information gleaned from omnipresent surveillance to predict future terrorist attacks.  However, Finch discovered that the computer was predicting ordinary crimes as well.  The government is not interested in these results, but Finch is determined to stop the predicted crimes.

Similar to the capability about which Golovin had written, and taking a page from the Person of Interest show referenced above, there are numerous predictive models being introduced in all phases of our everyday lives.  For example, Rick Shaw from has himself created a version of the system that Harold Finch created in an effort to stop bullying as well as school violence.

Known has TIPS, Shaw’s central repository system leverages a predictive analysis model to identify potential issues before they happen.  Based upon a scientific grid, Shaw claims that TIPS could have likely predicted (and prevented) tragedies such as the Virginia Tech shooting.

The key point is that these models or systems are based upon a capture, correlate and analyze process that is dependent on high speed technological interaction spanning many diverse points of data capture.  In this context the future is indeed now, as we do have this sharing or cross-pollination capability.  The obstacles to realization therefore are not technological, but are the result of the willingness (or unwillingness) on the part of independent stakeholders to truly come to together to share the required intelligence.

Once the relational elements of a collaborative model involving all stakeholders have been established, the not too distant future to which Golovin refers and a TV series suggests becomes reality today.

So here is the question . . . how strong are your organization’s collaborative links with its partners today?  How could the predictive model be used to improve supply chain performance?

In Part 2, I will examine more closely the application of this technological prowess in the world of procurement.  Meanwhile, be sure to check out Person of Interest on Tuesday’s at 10:00 PM EST on CBS.


Posted in: Commentary