An applied mathematics and statistics senior in the Whiting School of Engineering is using data to uncover the hidden systems behind one of the fastest-growing forms of human exploitation: online scam trafficking.
Hope Ugwuoke, whose interest in anti-trafficking efforts began in high school, was looking for a way to combine that passion with her data science studies at Johns Hopkins when she discovered that her advisor, Benjamin Huynh, an assistant professor in the Bloomberg School of Public Health, shared this interest.
Under the mentorship of Huynh and Sergey Kushnarev, senior lecturer in the Department of Applied Mathematics and Statistics, Ugwuoke set out to better understand the rise of scam compounds—locations where individuals are trafficked under false job promises and forced to carry out online scams—in Southeast Asia. She will present her findings at JHU Engineering Design Day, an annual event where students present their capstone projects; solutions that address real-world challenges.
While there’s been more attention paid to the increase in scam compounds over the past few years, the conditions that have allowed this growth remain poorly understood.
“It’s a relatively new problem, and there just isn’t a lot of research yet,” Ugwuoke says.
Ugwuoke decided to focus on three key forces often linked to scam compound growth: deforestation, weak governance, and armed conflict. Her goal was to identify connections between these forces through data analysis.
She started by collecting and layering data gathered from a wide range of sources, including United Nations maps of the locations of known scam compounds, satellite-based deforestation data, and data from the Armed Conflict Location & Event Data Project, which tracks incidents of violence at precise coordinates. Then, she began looking for clues. “I combined these coordinates to find patterns that could illustrate relationships between the variables,” Ugwuoke says.
Her analysis methods included combining visual mapping with statistical modeling. She used regression—a method for identifying relationships in data—to compare patterns across locations and applied a causal approach to explore whether scam compounds may directly contribute to deforestation.
“Because it is difficult to quantify governance, I looked at related indicators like conflict levels and access to basic services as a way to better understand where oversight might be weaker,” she says.
While the results were more nuanced than she expected, it was clear that governance was a key factor.
“Lack of governance is a big part of it,” she says. “When government systems are weak, it creates space for things like scam compounds to grow.”
The link between conflict and scam compounds was less direct than she had anticipated. Rather than occurring in the same locations, conflict appeared to have a broader effect. “Conflict in one area can weaken the entire region,” she says. “And that instability can allow scam compounds to develop elsewhere.”
Her analysis also revealed that scam compounds are often located in rural areas near national borders, where oversight may be limited.
These findings, which offer insight for policymakers and organizations working to combat trafficking, suggest that targeting only the compounds may not be enough.
“You don’t just want the areas with scam compounds to be stable, you need the entire region to be stable,” Ugwuoke says.
Another of her findings was that scam trafficking does not emerge in isolation.
“There’s always a combination of factors—location, governance, conflict—that allow something like this to grow,” she says.
Her research also indicates that scam compounds are not a primary driver of deforestation.
The project also presented her with a number of challenges, including that she sometimes had to rely on limited and delayed data. When conflict data is not released until up to a year after the events occur, it can affect the accuracy of her findings in that region.
Another challenge was that because the data she used came from a wide variety of sources, it was not presented in consistent ways, making it more challenging for her to identify patterns.
Even so, the work Ugwuoke did combining and visualizing the data sets is an important step toward a deeper understanding of how conditions may combine to create an environment in which scam compounds can thrive. By bringing together data that is often spread across different sources, she was able to discern patterns otherwise may have gone unnoticed, and her interpretations may help researchers and policymakers move beyond simply identifying connections to understanding how these factors influence each other.
“If we can identify these conditions, we can potentially change them and help put a stop to these dangerous practices,” she says.