How did you first get involved with START?
I first became aware of START as an undergraduate student while helping my professor conduct research for a book chapter. We used the Global Terrorism Database (GTD) to examine civilian deaths and injuries caused by al-Qaeda and its regional affiliates around the world. I officially joined START’s GTD team in November 2021, where I reviewed news articles to identify terrorist attacks and conducted quality control for the dataset.
What drew you to studying terrorism/radicalization (and using data-driven methods)?
As an undergraduate, I took a course called The Ethics of Counterterrorism, which sparked my interest in the field. Shortly afterward, I became a research assistant for Professor Avery Plaw, where my work focused on drone warfare and targeted killings.
My technical skill set developed largely out of necessity. I was involved in a project that did not have funding to bring on a computer scientist, so someone on the team had to learn how to code. Once I learned how to work with databases and build websites, creating visualizations and tools to analyze data became a natural next step.
In one sentence, what is your research focus?
My research focuses on drone proliferation, conflict and political violence data, natural language processing, and event data extraction and analysis.
What are you working on right now?
I am currently working on the Terrorism and Targeted Violence (T2V) in the United States project.
What’s the most interesting project you’ve worked on at START?
Building the Data Management System (DMS) for the T2V project and evaluating the potential efficiency gains of incorporating artificial intelligence (AI) into the production of political violence event datasets.
This work has given me a deeper understanding of every stage of the data collection workflow, from document classification and clustering to event identification and validation, and has helped me better understand where AI can meaningfully improve efficiency and, just as importantly, where human expertise remains essential.
What research gap are you most interested in tackling next?
There is still significant room for improvement in automated document classification and event extraction for data collection. Real-world documents are often messy; for example, news digests frequently contain multiple unrelated events within a single article. Developing methods to accurately segment these documents into discrete event sections before classification and clustering could substantially improve overall accuracy.
Another challenge is cross-document coreference resolution, where information about a single event is scattered across multiple sources that each provide only partial or complementary details. Improving document segmentation and cross-document event linking could help reduce the time lag between real-world events and data collection while also lowering the costs associated with producing event datasets.
How do you hope your research will be used in the real world?
There is growing demand for data that is accurate, reliable, and up to date, particularly in the area of political violence. I hope my work contributes to improving the way these data are collected, maintained, and analyzed, making them more useful for researchers, policymakers, and practitioners.
What do you do for fun outside of work?
I enjoy making music.