We’ve been told that we are special ever since we were little, but how special are we? But why is our brand experience often identical to everyone else’s? Somehow, in the eyes of big brands, we are just like everyone else.
After a deep dive into the inner workings of one class of prescriptive analytics—recommender system (personalization engines), it’s time to step back and explore how we can leverage prescriptive analytics in general. Today, I am going to outline 3 relevant use cases in business.
Each of us is pretty unique. The reason that some of us may appear similar to a brand is because brands typically don’t have enough data to tell us apart. Big data changes that fact! With social and behavior data, we will have enough data to view a user in over hundreds of different dimensions. With so many dimensions, the chance of finding 2 matching individual along all these hundreds of different dimensions is highly improbable. We are truly unique!
As the Charles Dickens novel goes “It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness”. This reminds me of the beginning of personalization and the challenges of the early approaches. Today, we live in a world where access to information is at our fingertip. In this world, relevance instead of raw impressions is more critical to the consumption of content, because an irrelevant impression is often ignored even if it’s right in front of us.
There is certainly a lot of exciting new developments around data science, especially with deep learning and artificial intelligence (AI). However, data science is actually not a new discipline in business. It was simply disguised under a different http://panelpicker.sxsw.com/vote/67189name—analytics. I want to revisit this topic of analytics because it is one of the most nebulous topics in the industry.
I thought I’d ease into this more technical subject by answering a question that I get asked many times: “how did you end up as a social media data scientist from your biophysics PhD background?”
Retrospectively, I have literally answered this question (in one form or another) over 100 times, with journalist/blogger interviews, in keynotes Q&As, or just casual conversations with colleagues or acquaintances.
Alright, now that you know my thoughts and feelings about the Klout acquisition, it's time to get back to some serious business—big data.
In my previous big data post, we discussed the minimum condition for any analytics to be actionable, which is summarized nicely by the actionability inequality—predictive window > reaction time. This condition is minimal in the sense that for any analytic to be actionable, it must at least satisfy this condition. It is a necessary but not a sufficient criterion for actionability. Therefore, to build truly actionable analytics, the actionability inequality alone is not sufficient. So what does it take to create actionable analytics? That will be the topics of our discussion today.
It’s been a while since we have talked about big data and analytics. So it’s time to come back to this topic. Previously, we talked about the need for sophisticated analytics to reduce petabytes of big data down to actionable bits. And there are 3 classes of analytics that could help us with this data reduction process:
Descriptive: Computing descriptive statistics that summarizes the existing data
Predictive: Building a predictive model and validating it, so it can be used to forecast data that doesn’t exist yet
Prescriptive: Building an actionable model with feedback to guide the decision maker to the desired outcome
But what is an actionable model? Many analytics vendors claim to provide actionable insights or actionable intelligence, but what do they really mean? What precisely are actionable analytics, and what make an insight actionable? These are the questions we will address today.