Data science is not a new concept by any stretch of the imagination. However, new technologies are making it easier for data scientists to capture and process huge amounts of information, which ultimately helps to provide better results from predictive measures.
To find out how data science has evolved through the years and how it is helping companies provide better outcomes for their clients and improve efficiencies internally, Peninsula Chronicle recently spoke with Keith McGhee, President of ITA International (ITA) Data Solutions in Newport News. ITA provides global service to customers, “In The Arena,” by leveraging subject matter expertise, data analytics, and technology to make what may seem impossible possible.
Peninsula Chronicle: What exactly is data science and how is it used?
Keith McGhee: Data science, when totally boiled down to the bare minimum, is really about making better decisions. We call it informed decisions. Informed by data. That takes on a whole range of possibilities. On the low end, just visualizing things, rather than looking at a spreadsheet of numbers, looking at a graph, trend line, that sort of thing. That’s data science. That’s helping people make more informed decisions by looking at trends over time. And then as you move up that value chain, when everybody thinks of data science, they think of machine learning or artificial intelligence, and some of the more number-intensive or algorithm-intensive applications. That certainly is a part of it as well. I like to think of it as a whole continuum of tools and techniques that can be used to help people make better decisions.
PC: Data science sounds like it has been around a while. It’s just getting a new face on it now.
KM: A lot of the techniques have been around, some of them for hundreds of years. It’s not that we just figured out how to do the math over the past two or three years. The math has been around for a while. Data science techniques have been used in statistical control processes in plants. W. Edwards Deming, I don’t know if you know that name, but he was a manufacturing guru when it came to thinking about how to arrange shop floors and things like that. Back in the 1950s or 1960s, they were doing different types of statistical processes. What has changed now, what’s really brought it into the mainstream, computing strength has gotten a lot more powerful. It’s gotten much more affordable to get computing power through a cloud provider, whereas that might not have been as easily accessible to smaller scale types of organizations as it is now. And just the vast amounts of data that has sprung up over the last five to ten years. There has been an extreme amount of data collection through YouTube, Facebook, all these places. Data is being collected every time we’re doing anything these days, such as with credit card transactions and all that. So, you put all that together, that’s really what has enabled the algorithms and the math that has been around for a while, to now come first and foremost.
PC: So, data science can be used in a variety of fields?
KM: Right. If you go onto Facebook or Google and you’re always getting hit with ads, that’s exactly what’s being done behind the scenes. They are collecting demographic data and they’re targeting advertisements toward people based on that. Other things that folks might not realize AI is a part of machine learning. So, if you’re using Google Maps, Waze, or any of those map apps, they are being driven by algorithms behind the scenes, pointing out the best ways to get from Point A to Point B, given traffic conditions and other data. So, it’s really everywhere, even though you might not think about it. When we talk to Alexa, there’s natural language processing that’s going on from the machine to turn your words into characters the machine can understand.
Here at ITA, we really try to practice what we preach. We’re a data-driven company. We’re always trying to learn, trying to get better. Within our company, we have a lot of what we call “dashboards.” They are ways to represent data to each other as well as everyone in the company. It’s really come a long way since I first started here at ITA and first started to work back a million years ago. We have it down to where each one of our departments has a dashboard that we consolidate information about our particular department and then present it, really via any medium. You can get it on your phone, you can get it on your computer, you can get it on your tablet. We make heavy use of that. Automation, data analytics to automate different processes and things within the company. That’s a very high priority for us. That whole-learning organization is critical for any company that wants to be data driven. They need to embrace the unknown and try to lean forward to figure out where technology and data science can help you, then work on trying to apply that toward your problems.
PC: If you were a high school or college guidance counselor and someone came to you saying they were interested in data science, what advice would you give them?
KM: There are two things there. Academically, I would point them toward statistics, and depending on how deep they wanted to go from a technical perspective, I’d definitely send them down an academic, technical path. You might not have to go as far as computer science or engineering, but something that’s going to get your hands in the mix so you can understand the technical piece of it. I think that would be critical.
If you move beyond high school and into the college level, encourage that student to get some kind of internship. Students can come into an organization and get embedded with a data analytics team. We do that at ITA every year. I’ll have one or two interns that will be part of my group. They’ll report here and work side by side with folks that are actually solving a customer’s problems. I hope they’re telling me the truth, but they tell me it’s a really good experience and they walk away with something they can apply. That’s my goal, to instill in them some things they can put on their resume, some problems that they tackled that have basis in the real world. All of them will have projects they’ve done in college, but when you hit the real world, things are maybe a little different than what you’re taught in college, so encouraging them to get that real-world experience through an internship I think would set them up: a) To get a job out of college when you’re looking for that first full-time job or first professional job after college, but also b) It’ll give them some skills they can carry along with them throughout their career. And if it works out, I’ve hired interns here, too. A lot of times, if you do a good job with your internship, it’ll help you get that first job because you’ll just come right back and pick up where you left off.
PC: Sounds like an internship might also be a good way to find out if data science is not the path for certain people.
KM: Right. I would contend, if you head off down one of these paths with the technical side of it, you’ll learn skills. And if it’s not right for you, you’ll still have learned something, and I think that can help you later in life.