Predictive Analytics is the buzzword on the lips of CIO’s returning from Big Data conferences this year. They have heard amazing stories, some similar to this one:
Imagine if you will a world in which your phone checks your calendar in the middle of the night and notes that you’ve got an early meeting. Pulling information from your car’s GPS, it determines that you will need extra time to get to work, since your usual route has construction starting today. So it instructs your alarm clock to wakes you up no less than fifteen minutes early, but still at the point of your lightest sleep. Your smart-watch makes the final decision as to when to vibrate and wake you. Your house knows based on your shower times and water usage that you will shower within three minutes of your wake-up alarm and about an hour before you leave your house. The coffee-maker starts automatically when you’re halfway through your shower, and your refrigerator senses that your milk is about to expire, so it puts a quart on your shopping list. And all this occurs before you’re even fully dressed.
Although I’ve made this example up, all of these technological wonders are either available in the present day or nearly-so. They’re examples of the power of predictive analytics that we can imagine in the future as “the Internet of things” allows us to understand ourselves and our interactions with the world more completely and deeply.
So why don’t we have this kind of predictive analytics, and what does this mean for us as we talk to our CIO’s about the implications for our businesses? The missing ingredient in this formula, the one that sits squarely between the reality of our smart phones, wearable tech, and home automation suites, is the integration of these data streams and translation into interconnected and digestible data volumes. Right now my car’s GPS can’t talk to my home network to see the traffic alert so that it can tell my alarm clock to wake me up early, and then confirm with my smart watch (via its built-in accelerometer) to make sure I’m in a light phase of my sleep when it does so. And (perhaps the greatest tragedy of all), none of these talk to my coffee maker to have my morning java ready fifteen minutes early!
Data architects and data scientists who tangle with the rich and fertile world of predictive analytics in the coming months and years will struggle most with the integration of data streams to provide the most enriching possible set of analysis paths. The even bigger challenge however is this: what makes sense to us now in terms of the ways we can combine disparate streams of data to provide meaningful insights into the world around us is likely not the same set of analytics combinations that will be of greatest value to us in the future. Likely we will find ourselves combining human intuition and mathematically-informed data science with computer-driven machine learning to discover unanticipated ways to combine data streams for even more insightful revelations about us and our world. Who knows what insights will emerge from those new kinds of analytical forays!
DataHub Writer: Douglas R. Briggs
Mr. Briggs has been active in the fields of Data Warehousing and Business Intelligence for the entirety of his 17-year career. He was responsible for the early adoption and promulgation of BI at one of the world’s largest consumer product companies and developed their initial BI competency centre. He has consulted with numerous other companies about effective BI practices. He holds a Master of Science degree in Computer Science from the University of Illinois at Urbana-Champaign and a Bachelor of Arts degree from Williams College (Mass)..
View Linkedin Profile->
Other Articles by Douglas->