Sluething Cyber Crime

Baylor computer science professors lead NSF-funded effort to identify illegal activity online

Many people use consumer-to-consumer websites for common household transactions. These sites, like Craigslist, connect buyers and sellers for a variety of legal transactions. Unfortunately, criminals likewise utilize these websites but do so to facilitate illicit business in human trafficking, the sale of stolen goods and more. It’s these types of transactions that two Baylor University professors and an interdisciplinary team of computer scientists are looking to thwart.

Pablo Rivas, Ph.D., assistant professor of computer science in Baylor’s School of Engineering and Computer Science, is the principal investigator on a grant from the National Science Foundation (NSF) to utilize technology to identify and disrupt illicit transactions online. Baylor colleague Tomas Cerny, Ph.D., assistant professor of computer science, also serves on the five-person team.

“We are looking at this from two perspectives,” Rivas said. “One is for human services being offered illegally, with the goal of detecting human trafficking. Second, what we learn from this domain can be applied to other transactions, like stolen goods such as automobile parts.”

The project sits at the intersection of emerging technologies and human challenges. The NSF funding will fuel the research team as they apply their discipline in a way that could serve individuals in need of an advocate.

Rivas, Cerny and their team will seek to determine the effectiveness of a technology called natural language processing (NLP) to identify suspicious listings online. NLP is considered a subfield of artificial intelligence (AI) involving human language patterns which pursues an understanding of language, context, information and more shared online.

“NLP has been around for a long time, but computational linguistics and increased computer power in combination with machine learning breakthroughs have pushed the field to new exciting frontiers,” Rivas said. “With machine learning, we can thrust NLP to make inferences by detecting patterns in language.”

The specific grant awarded to this team is given to projects with the potential for risk and reward. These projects are, by nature, experimental. However, if they work, they advance safety and security for internet users. The risk, Rivas said, is that researchers don’t know what they’ll find. But, the goal of disrupting illicit activity, identifying individuals caught in trafficking and making it harder to engage in such activity is a goal worthy of that investment.

In addition to Rivas and Cerny, a Baylor alumna and former Baylor professor serve on the grant team. Laurie Giddens, M.S.I.S. ’02, Ph.D. ’17, assistant professor of information systems at the University of North Texas, is a two-time Baylor graduate who partnered with her then-Baylor professor, Stacie Petter, Ph.D., now at Wake Forest University.

“We believe that, with machine learning, we can create models that can help us learn more about how crime works,” Rivas said. “We can then provide that intelligence to others, like law enforcement or behavioral scientists, to recognize those engaged in illegal activity and connect law enforcement to them, so they account for their doings. That leads to a safer online community for everyone using these consumer-to-consumer sites and disrupts illegal activity.”