On 23 April, representatives from 12 New Jersey universities known as the New Jersey Big Data Alliance met to discuss how to increase research in the area of big data and help drive economic development in the state.
The founding universities of the alliance, which was formed in December 2012, include the following: Kean, Montclair State, New Jersey Institute of Technology (NJIT), Rowan, Rutgers, Stevens Institute of Technology, Stockton, as well as affiliate member, Princeton University. The schools will work with government agencies and local industries to create training and career development workshops, a portal for big-data resources, and increase partnerships with others in the field.
Speakers at the symposium included representatives from Bell Labs, IBM, and McKinsey & Company, which offers business solutions to some of the top companies in the world. They covered several challenges, including a lack of storage capacity for the growing amount of raw data, the difficulty with developing meaningful information with analytics, and the shortage of big-data specialists. IEEE was a supporter of the event.
THE BIG CHALLENGE
Keynote speaker, Christopher Greer, spoke about the growing amount of data as more and more items become connected to the Internet, known as the Internet of Things. He is the director of the National Institute of Standards and Technology (NIST), based in Gaithersburg, Md. Greer cited that about 44 trillion gigabytes of data would be available by 2020, compared to 4.4 trillion GB today. This leads to concerns about a shortage in storage capacity, as well as a need for data cleansing. “How do we decide what to keep and what to let go of, and for how long?” he asked. One of the challenges would be to put mechanisms in place that can help determine what information should be deleted.
Moreover, Greer brought up the challenge of time-critical applications of data to inform decision-making. For example, with smarter cars, sensors that can determine when a pedestrian is too close to the vehicle uses analytics to also decide what action the car should take and at what time. “In these situations, data has to give us the acceptable answer at the right time; not a right answer at an acceptable time,” Greer said. As helpful as data might be to make better decisions, whether in business, health care, or transportation, “Good sources of data can also make poor decisions if we’re not careful,” he said.
In a panel discussion on big data challenges and opportunities, IEEE Member Parag Pruthi explained the dangers that can arise when interpreting data incorrectly. He is the founder and CEO of Niksun, a company that develops monitoring and cybersecurity systems, in Princeton, N.J. To make his point that correlation does not always equal causation, he presented this chart from Gizmodo that makes it appear as if a decrease in Internet Explorer users is the reason for a decrease in murder rates. On the same panel, Lisa Sokol, an analytics data scientist at IBM, said data analysis is not about testing a hypothesis, but rather “playing with the data and looking for the errors.”
In the “Preparing Tomorrow’s Workforce for the Big Data Economy” panel discussion, according to a report from the McKinsey Institute, there is a shortage of some 150 000 people with analytical skills in the United States alone, as well as a shortage of 1.5 million managers and analysts who can interpret data and implement results, Several of the universities have already formed programs to help address this problem, including NJIT, which launched a one-year certificate program on Big Data Management and Mining.
Several attendees asked the alliance about the potential to work on start-up efforts on the local scale, as well as having the state using big data for collective budgeting, in which citizens can see how funds are allocated to have a say over where tax money goes. This is currently being tried in several cities, including Oakland, Calif.
To stay updated, visit the New Jersey Big Data Alliance website: www.NJBigData.org. You can also join as a member, contribute to the big-data portal, or be a speaker at a future event.