You don't have to be a data scientist, but you do have to try and understand them. ASU's new program helps domain experts better communicate on big data projects.
Big data -- you either understand it or you don't.
"Some people perceive the big data methods as sort of like stirring up some voodoo," said Professor Michael Goul, chair of the Information Systems Department at the W. P. Carey School of Business at Arizona State University.
It's a problem. But the thing is, big data is not going away. According to the university, 85% of Fortune 500 companies have a big data initiative underway or in the works.
So, how can companies forge some understanding between their data scientists and domain experts?
ASU is hoping to help through the introduction of its new business analytics certificate.
It's not aimed at turning people into full blown data scientists, but rather, at enabling those who work with data scientists to better understand and communicate with them through a fundamental understanding of analytics in a business setting.
"If you're a manager saying 'I think this big data can make a big change in, say, how we do our marketing to this specific clientele that we have,' and you know that corporate is going to send you out some data scientists that you'll be working with to develop some new solutions -- you want to meet them more than half way," Goul said.
The programs lasts about six months, and includes four courses, all online. There are weekly quizzes and readings, but students have the ability to pace themselves.
Two of the classes are core classes. Business Analytics Strategy covers how businesses have realized benefits from business analytics, Goul said. Students will do cases studies on companies that invested in analytics, why they did it, and what they intended to get out of it, as well as any pitfalls they encountered. The other course is Foundations of Applied Analytics. Students learn about analytics models -- what they are, why they're valuable, if they decay over time, and how to prepare a good model. Also, what's important to communicate to data scientists about the model they're building.
From there, the program has two tracks. The first is in supply chain management -- Analytical Decision Modeling I and II. The other track is data management -- Enterprise Analytics and Big Data and Data Mining and Predictive Analytics.
Goul said there are a few reasons to foster an understanding of analytics.
"The domain expert probably has a solid understanding of the data that's being collected. They know why it's being collected, what purpose it's served in the past. They know things like 'Oh, this piece of data probably tells us the same information as this piece of data,' and understanding then how the methods that are going to be used by the data scientists could make use of that information is just golden," he said.
On the supply chain side, he gave an example of a shipping company collecting data on its trucks. They might know where every single one of their trucks is at any given time, so if something happens along a route, they can reroute the trucks.
"Being able to work with the data scientist to understand how that new information and what might be gleaned from it can be leveraged would be key," he said.
He said that having a partnership between people who understand a domain and people who understand the data is important these days.
"My hope is that this will provide people with the insights that will allow them to communicate on a different level and really address some nasty problems that we probably haven't looked at before," he said.