Daniel Bayer
   Staff Writer

Dr. Shih-Lung Shaw, professor of geology at the University of Tennessee at Knoxville, gave a public lecture last Friday, February 19, titled “Big Data, Human Dynamics and Space-Time GIS.”

Shaw began his lecture by noting to those in attendance that everything being done on one’s cell phone is constantly tracked and recorded.

Rather than encouraging hysteria, however, Shaw explained that the collection of such data can be extremely useful in areas such as urban planning and the scheduling of public transportation.

“How many of you have your cell phone with you?” asked Shaw at the beginning of his talk, “Big Data, Human Dynamics and Space-Time GIS.” “How many of you have Gmail? How many of you have Facebook? You and me… we’re all part of the big data.”

The “big data” that Dr. Shaw referred to is the immense gathering of information on what people do every day – calls, texts, purchases, travel. Using this information, behavior can be tracked to a remarkably accurate degree, says Shaw.

“We are being tracked all the time, and that data is collected by those companies, government agencies and many different entities,” says Shaw. Analytical programs can then determine a person’s movements, both in the real world and the virtual one.

“We are not random,” says Shaw of the patterns of human movements throughout the day. “Using cell phone data, we can develop a 93 percent potential predictability in user mobility. In other words, I can predict where you’ll be 93 percent of the time.”

The resulting “human dynamics” are of great interest to geologists, he says.

“For example, how do individuals in Greensboro move around everyday? Which parts of Greensboro share similar spatial-temporal movement patterns?” says Shaw.

“Spatial-Temporal” refers to the relationship between physical space and the passage of time, and how these two elements interact in the way individuals navigate their surroundings. The resulting patterns can be analyzed to allocate transportation resources and plan future land use, says Shaw.

“This can be very useful for planning, for transportation,” he says.

The study of these space-time patterns is called geographic information science, or GIS. The new challenge, says Shaw, is the development of the online, or “virtual” world.

“Physical space and virtual space are not separate and independent,” says Shaw.

Virtual space would be things like emails, cell phones and websites, and people’s interactions within this world requires a shift in thinking for those working in GIS, he says.

“What we do in virtual space influences our actions in physical space,” he says. “To give you an example. Ten years ago we used to go to Bestbuy for electronics. How many of you still go to Bestbuy? When you go to Bestbuy, what do you do? You touch it, you feel it, you try it, then you pull out your phone and find the best price online and order it.”

This new interaction changes the way that communities approached certain issues in the past, says Shaw. “This is going to change the needs for open space, our needs for parks, our needs for schools and so forth.”

It also affects GIS itself, he says.

“How can GIS handle e-shopping, online games, and tracking apps? How should we represent activities and interactions in virtual space and connect them to persons in physical space?”

Shaw then used examples from France and China to demonstrate the movement of people through space and time, and how the virtual world interacts with the physical one.

Using cell phone records, Shaw followed the movements of multiple individuals around the city of Paris throughout their day. Among the findings were that the daily routines of people differed based on their socio-economic status.

In China, the data compared the routines of people in two different cities, information that could be used to promote the development of alternative transportation, he says.

Despite these developments, questions still remain, says Shaw.

“How big is big enough?” when it comes to the amount of data gathered, asked Shaw. He finished his lecture by asking the open-ended question, “what kind of research questions are good for big data analytics?”