HP Big Data Blog
Big Data changes everything. The HP Big Data team is powering data analytics, information governance, and information management across the enterprise. Join the conversation.

Analytics for Human Information: Staying One Step Ahead of Predictive Analytics

I attended a Big Data event in Miami recently. While working at the booth and listening to presentations, it became clear that Predictive Analytics was the hot topic. IDOL seemed to intrigue people based on its ability to uncover breaking topics by making new clusters available. But there was some skepticism.


As I considered a host of questions from the audience and attendees, it was clear that they still had one foot in the old world and one in the new. They needed to shift their thinking.


predictive analytics.jpg


My first question went to a robust PhD, “What is predictive analytics?”


He looked at me quizzically and said, “It’s about understanding what your customer is going to do before he does it.”


“I see,” I responded, “So this must assume you already know all the permutations of what this customer will do? Hmm, what happens when he does something you did not predict?”


“Well, we find it,” he replied.


I asked, “Sounds tough, how?”


“Well we look for breaking trends,” he said.


“How is breaking defined?” I asked.


“As topics that break a set threshold of popularity,” he replied.


“What if they do not break a threshold, but stay underneath for some time based on market size or some other variable?” I persisted.


“Well I am not sure what you mean…silence…” was his response.


“So, it sounds as if you are stuck in the comfort zone of what you already know, and then you use historical data and trends to inform what will happen in the future—knowing full well that past results are in no way an indicator of what will happen next.” I explained.


“Yes, but, well, I am not sure how else to do it…more silence…” he replied.


And with my final comment, I had just the lead-in needed to get to the next step, “The key is you don’t have to—IDOL does it for you by using clustering technology to reveal breaking concepts across a corpus of data. Want to see a demo?”




Edited by Robin Hardy

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the Community Guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Showing results for 
Search instead for 
Do you mean 
About the Author
Dan Burke serves as a vice president of HP Software. He is in SME within Power and Emerging technologies, focused on HP Autonomy Explore and...

Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation.