One of the big questions right now is where society is going to find enough data scientists to fill the need. McKinsey & Co. predicted in 2011 that the U.S. would face a shortage of 140,000 to 190,000 people with deep analytical skills by 2018. The predictions haven’t improved since then.
That’s led some to advocate for a data team approach, but that doesn’t change the fact that we’re still falling short in terms of data and analytics literacy.
Recently, I stumbled across Vincent Granville’s “Four Easy Steps to Becoming a Data Scientist.” Of course, I had to read it because it was the first time I’d ever encountered anything suggesting that there might only be four steps, let alone easy ones, to becoming a data scientist.
Granville’s bio says he has worked with Big Data, predictive modeling and business analytics for the past 15 years. That includes working on real-time credit card fraud detection with Visa and advertising mix optimization with CNET, automated bidding with eBay and other high-level projects.
Given how much time and work he’s put into data science, I began to suspect that “easy” might be relative — and, completely unrelated to time.
First, Granville recommends you buy a “modern data science book.” The key here is to steer clear of statistics textbooks; he recommends you go ahead and start with his book, “Developing Analytic Talent: Becoming a Data Scientist” or wait for his new book, which releases in April. Okay, that’s a bit self-serving, but I have to admit, at 336 pages, we’re still in “easy” territory.
Second, he offers links to cheat sheets available on Data Science Central, which he launched. It’s a community site and definitely a stop for Big Data reading. Heck, that’s how I found his post, right? Here’s the thing about this step, though: At this point, he’s sending you on a research journey that includes studying Python, machine learning, deep learning, Hadoop, R programming and more. Already, we’re pushing the envelope of “easy.”
By step three, I knew he was pulling my leg with this easy stuff. At this point, you move on to working with real data, preferably by enrolling in an apprenticeship or enrolling in a specialty program which, he admits, might require a Ph.D. Step four is, “Launch your career” and includes links to a job site, advice about starting your own consultancy or start-up, etc.
While those are technically four steps, none strike me as particularly easy, unless you have an advanced mathematics degree or years of specialized research and analytics work behind you already.
Still, I will give him this: It is certainly much easier than the lengthy and multiple lists in the “Big-Data-Enabled Specialist” skills profile recently released by the Oceans of Data Institute.
A panel of experts from Google, Microsoft, Columbia University, George Mason University and NASA, among others, sat down earlier this year to produce this profile. Then, 150 big data professionals reviewed and added their opinions through an online survey.
The end goal is not to cultivate data scientists from the existing workplace, but rather to give educators, policy makers and those already in the profession a roadmap for the next generation of data scientists. That’s right, Millennials, your education is already outdated.
This is not a light profile. You can read it for yourself, but I think it’s safe to say that you’ll want to consider specializing by sixth grade, at the latest. It includes everything from time management and writing to visualization design, machine learning, parallel programming, research methodology and brute-force analytics.
There are separate lists for the tools and trends/corners you should know, plus another list for “five years from now.” And then there’s a table that couples duties with tasks in this “learning occupation.”
Of course, much of this would be covered in a robust computer science program — or two. Reading through the profile, it’s hard to imagine what primary and secondary schools might bring to the table, but that’s explored more in the executive summary:
“Unexpectedly, ‘soft skills’ such as analytical thinking, critical thinking, and problem solving dominated the 20+ big data skill and knowledge requirements identified by the panel and endorsed by experts who completed the validation survey.”
If you’re curious about who served on the panel, you’ll find the biographies here. You can also download the Big-Data-Enabled Specialist profile itself by providing basic user information.
Loraine Lawson is a veteran technology reporter and blogger. She currently writes the Integration blog for IT Business Edge, which covers all aspects of integration technology, including data governance and best practices. She has also covered IT/Business Alignment and IT Security for IT Business Edge. Before becoming a freelance writer, Lawson worked at TechRepublic as a site editor and writer, covering mobile, IT management, IT security and other technology trends. Previously, she was a webmaster at the Kentucky Transportation Cabinet and a newspaper journalist. Follow Lawson at Google+ and on Twitter.