1:02 p.m., Nov. 8, 2012–Every minute of every day, data grows exponentially. Web and social media activities, cell phone and GPS signals, business transaction records, digital documents and more generate massive amounts of data. In fact, this infographic by Josh James, founder and CEO of the business intelligence company Domo, shows Facebook users share 684,478 pieces of content, Google receives over two million search queries and consumers spend $272,070 on web shopping each minute.
And while these numbers are almost inconceivable to the human mind, this big data is not only measureable but holds important implications for businesses of all sizes and types.
More than 50 faculty members and industry partners met to talk about these implications of big data on financial services, pharmaceuticals and education at a conference held in late October at Clayton Hall on the University of Delaware’s Laird Campus.
Sponsored by UD’s Institute for Financial Services Analytics (IFSA) in the Alfred Lerner College of Business and Economics and JPMorgan Chase, the conference featured presentations by experts from JPMorgan Chase, Teradata Corp., AstraZeneca Pharmaceuticals and SevOne Corp., as well as a panel of faculty members from the Lerner College and the College of Engineering.
UD President Patrick Harker told attendees the collaboration between JPMorgan Chase and the IFSA is at the intersection of industry and education and noted the benefits go both ways, from informing industry operations and efficiencies to revolutionizing teaching and shaping UD’s business and engineering curricula.
“This collaboration with JPMorgan Chase — our work in guiding a continually transforming industry and preparing technologists for it — has been one of the most rewarding for us,” said Harker. “I know the IFSA will launch some incredible innovations in the financial services sectors and of course, managing big data is central to this outcome: capturing it, storing it, searching, sharing, analyzing, visualizing it and, ultimately, exploiting it.”
Raghav Madhavan, managing director at JPMorgan Chase, began the conference with an explanation of the “three V’s” that characterize big data – volume, velocity and variety – and noted that transactions, interactions and observations define this mode of understanding and analyzing data.
“Data science is often cast in a mystical kind of way,” said Madhavan. “Big data means learning from data of all types, focusing on scale not sample, applying multiple disciplines like business and engineering, and the rigor of validated findings all focused on a business goal.”
Madhavan shared James’ infographic to make the point that dealing with massive amounts of data, storage and transactions are essential to business planning and decision-making.
“Think about the impact of taking just 10 percent of this information and being able to process and analyze it,” said Madhavan. “From a social media context, we can figure out what is being said, why it is relevant to us and how we can monetize it so that it is beneficial to us. For example, how do we understand the genuine concerns people have when they Tweet with a certain hash tag?”
David Schrader, director of strategy and marketing with Teradata Corp., called the processing of emerging data types the really “interesting piece” in his presentation, “Big Data for the Financial Industry: New Insights, New Value?!”
“As someone is using an ATM are they smiling, frowning?” asked Schrader. “This is about providing an environment so no matter what structure data comes in, you are able to process it, extract the insight and connect the dots back to the customer record.”
Schrader shared his company’s CSI-style education YouTube videos on “StagnoBank” to show the capabilities of big data.
“These videos show us what kinds of new insights are possible,” said Schrader. “We can use analytics to dig for better marketing, better customer service and better mobile apps.”
Shifting gears, Anastasia Christianson, senior director and global discipline leader for biomedical informatics with AstraZeneca Pharmaceuticals, spoke about “Getting More Knowledge Out of Our Data for Better Decisions.”
By focusing on drug discovery and development, she explained how big data offers pharmaceuticals and health care the opportunity to combine data and information in novel ways.
“With larger volumes of data generated stored in heterogenous formats, scientists are spending more time organizing data than analyzing it,” said Christianson. “Scientists are making critical decisions on knowledge gained from only a fraction of data and information available so the question becomes, how can we make the best use of the data available to us?”
According to Christianson, four areas of great opportunities include translational medicine, or translating preclinical science into patience studies; predictive science, or applying modeling and simulation techniques in pharmaceutical research and development; and personalized health care, or delivering the right treatment to the right patient in the right dose at the right time.
Real world evidence, the fourth area, means taking information outside of controlled trials to create insights on diseases, products and patient populations.
“This is data like those found in claims databases, patient diaries, electronic medical records,” said Christianson. “Real world evidence is critical to our ability to deliver on our strategic goals. The effects of better use of big data then are better characterization of diseases and patient populations; development of new products and therapies and assessment of those already in use; and better targeted products.”
Using social media as an example, Christianson explained a lot can be learned from what patients are saying.
“Sixty-four percent of adults with chronic disease go online and 81 percent of adults without chronic diseases go online,” said Christianson, noting those adults make comments in blogs, social media and other outlets. “We need to create ways to best analyze that data to get the real nuggets versus regular chatter. Khalil Gibran put it best: a little knowledge that acts is worth infinitely more than much knowledge that is idle.”
Vess Bakalov, chief technology officer and senior vice president of products with SevOne Corp., gave the final presentation on “How Big Data Helps Manage Big IT Infrastructure? Making Sense of Machine Chatter.”
“I like to add a fourth ‘V’ to the mix—veracity—when we talk about big data,” said Bakalov, who is also a UD alumnus. “How real is the data? In many ways, the inferences we make from the data need to be based on something real. You can’t act on data you don’t trust.”
Bakalov also focused more on big data as it relates to machines.
“Even when you are dealing with people, the nature of data comes from machines,” said Bakalov. “We need to talk about the rise of the machines. Thousands of devices generate millions of metrics, things like temperature, bandwidth, power. And the size of the data is different: the HVAC system used in this building is generating completely different particles from the supply logistics system for the hotel up the road.”
For Bakalov and SevOne Corp., then, creating an open architecture that can take all types of data from across the world and present the user with a single view is a key piece to understanding big data.
The conference concluded with a panel session moderated by Bintong Chen, professor of business administration, and featuring Hui Fang, assistant professor of computer and electrical engineering; Harry Wang, associate professor of accounting and information systems; Skip White, professor of accounting and information systems; and Cathy Wu, Edward G. Jefferson Chair of Bioinformatics and Computational Biology.
Article by Kathryn Meier