Articles

Latest Trends in Artificial Intelligence, Cybersecurity and the Dark Web Revealed at the Florida Business Analytics Forum

By Keith Morelli

Florida Business Analytics Forum

TAMPA (May 14, 2019) -- When you are unlucky enough to find out a stranger has already received your tax refund, or that you purchased a new living room suite in California when you’ve never been to California, your first thought is which restaurant server or convenience store clerk had your credit card last night.

Yes, someone did steal your identity, but likely it wasn’t recent.

The real awareness among the general public that identity theft was a “thing” came about five years ago when Target and a slew of other smaller companies announced that more than a billion records had been breached.

“That was a milestone year for cybersecurity,” said Hsinchun Chen, director of the University of Arizona’s Artificial Intelligence Lab. Chen was one of three key presenters at the third annual Florida Business Analytics Forum held Tuesday in the Marshall Student Center’s ballroom. Nearly 400 people attended to learn about the latest trends in data analytics.

Chen talked about recent discoveries concerning the dark web and developments to try to protect identities of people and proprietary information of businesses and corporations. Though Target announced the discovery of its breach in 2014, he said, the hacking could have been going on for years. Not only was customer personal information compromised, he said, but Target saw a 46 percent drop in profits that year.

Now, artificial intelligence is being enlisted to dive into the dark web to come up with solutions and determine who is behind the thievery and other illegal activities ranging from child pornography to terrorist activities. Chen had worked with the government several years ago in uncovering digital platforms used by jihadists and came up with algorithms that, based on language and other factors, were able to narrow down which terrorist group – or person – was online and what they or he were saying.

Besides Chen, experts in the fields of mobile analytics and artificial intelligence in the criminal justice fields also delivered speeches and participated in panel discussions on topics ranging from dark-web threats to fake news.

The forum, hosted by the Center for Analytics and Creativity at the University of South Florida’s Muma College of Business and is sponsored by the SunTrust Foundation, was designed to lead to real changes in the way businesses implement analytics. It’s what motivated the speakers to attend the event, said Balaji Padmanabhan, director of the Center for Analytics and Creativity.

“You cannot get these individuals to come for any amount of money,” he said. “They have graciously and generously given us a big chunk of their time. What we promised them in return was impact, that their ideas will set things in motion in our community and that we will translate some of their ideas into practice to drive the next wave of analytics innovations in Florida.

“We want you to learn today,” he said, “and implement tomorrow.”

Hari Balakrsishnan, a professor of computer science at the Massachusetts Institute of Technology, has conducted research into networked computer systems and came up with a mobile analytics system that can measure how well a person drives. The system is used by insurance companies and mobile fleet supervisors to determine risks. It also can be used to reward good drivers, he told the attendees, and some insurance agencies are beginning to reward customers who display safe driving habits, he said.

“Twenty-five percent of people,” he said, “want insurance rates based on how they drive.”

The DriveWell tag is a small device that sticks to the windshield and is linked to the driver’s cell phone. It can sense when a car brakes suddenly or accelerates to a high speed. Balakrishnan said he has one on his daughter’s vehicle and can see how she drives to school each day.

“After seeing her driving score,” he said, “I can tell you she’s not the worst driver in our family. And that’s based on data.”

The goal is to make people aware of bad driving habits so that those habits go away and the roads are safer for everyone. 

The third main speaker, Alex Chohlas-Wood, is now the deputy director of the Stanford Computational Policy Lab at Stanford University, but he had been the director of analytics with the New York Police Department, where he developed software to track criminal behavior across the entire city.

Data, he said, has been collected by the city for decades, but only recently has it been readily available to detectives. He said he formulated Patternizr, a software that groups similar crimes, to allow police to see trends that may not have been apparent before.

“A detective cannot read reports from all the precincts,” he said. This information now is available “in nanoseconds.”

More recently, he and a group at Stanford, have begun work on a system aimed at reducing inmate populations at county jails. The group is looking at the number of people arrested who spend two to five days in jail before they are released because prosecutors decide to drop the charges. A third of those who are released spend more than three days in jail.

His algorithm examines all aspects of the crime and police reports and determines the probability of a suspect’s case being dismissed. That results in those people being released early, freeing up space for more hardened criminals.

The forum also featured two panel discussions, the first addressing analytics and fake news. The panelists include Eduardo Hauser, managing director of Discovery in Miami, and Josie Hollingsworth, audience engagement editor at Politifact in St. Petersburg. The second panel discuss healthcare analytics and featured Paul Kuo, head of surgery at Tampa General Hospital and Kedar Kulkarni, director of data science at Florida Blue.