Marketing & Communications

Making sense of 'big data'


Business Analytics makes sense of 'big data'

In a now nearly iconic story, a father is outraged when he finds coupons for pregnancy-related items sent to his teenage daughter by Target, only to discover that she is indeed pregnant. How did the giant retailer know such an intimate detail about her before her own father did? The answer: Data mining.

The story, widely reported in the New York Times and elsewhere, highlights a growing field called business analytics that can help guide companies' marketing decisions and business practices based on information they collect - from people's Internet use and shopping history and from myriad other sources.

"We know that Google and other Internet-based businesses such as Facebook depend on information for their ads; data are really their essence," said Dr. Arthur Yeh, chair of the applied statistics and operations research department. "But traditional companies like Cooper Tire have been collecting customer data for many years as well. The question is what to do with these data."

In response, the College of Business Administration has for two years been offering a specialization in business analytics and intelligence, which Yeh characterizes as a "marriage among statistics, operations research, computer science and management of information systems."

Trevor Bischoff, a senior from Napoleon triple majoring in accounting, finance and business analytics and intelligence, said, "What's most appealing about this specialization is that it combines everything else you're learning and helps you derive meaning from it."

By drawing on each of these disciplines, data miners gain new insight. Dr. Christopher Rump, applied statistics and operations research, directs the specialization and teaches the required data mining class in applied statistics. He explained that the "intelligence" part of the program's name refers to the ability to not only interpret data but to communicate their meaning.

"What's most appealing about this specialization is that it combines everything else you're learning and helps you derive meaning from it."Bischoff especially enjoys that challenge. "The data mining and the business intelligence are parallel. It's not only deriving the meaning but being able to convey it perfectly. What do you want to say? What is the best way to display it? We use a number of software applications that allow us to present it in different ways. We can create a sort of 'dashboard' to display trends and key performance metrics, and see what will happen if we vary a factor."

This helps businesses make better predictions based on the sea of data they are collecting, Rump said.

Reaching out to the business community, the college will host a symposium on best practices in business analytics on April 26 featuring a variety of business and academic experts speaking about analytics practices and applications in diverse industries.

In addition, the University is in the last stages of planning and approval for a one-year, full-time, interdisciplinary master's degree in business analytics uniting the accounting and management information systems and applied statistics and operations research departments from the College of Business with the computer science and mathematics and statistics departments from the College of Arts and Sciences. Along with Yeh, Drs. Hanfeng Chen, mathematic and statistics, and Joe Chao, computer science, are collaborating on developing the new graduate program, which will be administered by the Graduate College.

Currently, 28 students are enrolled in the undergraduate specialization, which can enhance any major and should make graduates with those analytical skills more appealing to potential employers, according to Yeh.

"The ability to analyze data will be helpful in any area of business, from finance to accounting to supply chain management," he said. "In fact, it's almost becoming a must. We expect in five to 10 years that the vast majority of jobs for business students will require analytics skills.

"When we set out to do this it was in response to a trend. As more and more business jump on the Big Data bandwagon, we envision increasing demands on the job market for people with analytical skills," he said.

Students pursuing a bachelor of science degree in business administration may add the specialization by taking two additional statistics courses and two management of information systems courses geared toward analytics. There is no additional math required, although the field appeals to people who like dealing with numbers and have an aptitude for quantitative analysis, Yeh said.

"You have to have a sense of curiosity to enjoy this," Bischoff said. "You have to be curious about the natural trends in everyday life. It's as if this gives us a big spreadsheet with data and teaches us how to explain it to someone else. We have a saying that data don't talk to strangers. You have to get to know your data."

"What is powerful about this is that it is data-driven and fact-based," Yeh said. "While we can never throw out intuition, human brains cannot process the huge amounts of data that are available today very inexpensively. But numbers can yield some surprises, and companies can make good use of the information if they have employees with the skills to analyze it."

"As more and more business jump on the Big Data bandwagon, we envision increasing demands on the job market for people with analytical skills""Every day there are more examples of the impact of business analytics," Rump added, noting a recent article in Business Week about Coca-Cola utilizing 600 flavor variables to create a formula for standardized orange juice. "Companies are innovating and using data in ways we haven't seen before," he said.

Another example is Google Flu Trends, which in some cases more correctly predicted the severity of a year's flu season than the Centers for Disease Control and Prevention. Then there is New York Times reporter Nate Silver, who correctly predicted the winners by state in the 2012 presidential elections. "He's a rock star" in data analytics, Yeh said.

On the other end of the election process, Yeh pointed to Barack Obama's team of campaign strategists, who used analytics to determine where best to place their resources and when. "I think a lot of the credit for his election is due to them," he said.

"Data collection has been happening for a long time, whether we like it or not. From a business perspective, it is better to base your decisions on data than intuition alone."

 (Posted February 25, 2013 )