Student in Focus: Fiona Jiang, Scientist


Fiona Jiang

Above is a visualization of Fiona’s neural network, titled “Figure 9. t-SNE Visualization of Node Embeddings Generated by GraphSAGE.”This graph shows the relationships between the data on innovation Fiona collected and analyzed.

Webb senior, dorm prefect, and Filmmaking Club president Fiona Jiang (’22) was one of just 300 students to be named a Regeneron Science Talent Search Scholar this year. Her winning project? A neural network tracking patterns in innovation using data from the US Patent Office, “Characterizing Decades of Technological Advances with Graph Neural Networks: An Innovation Network Perspective.”

An artificial neural network is a set of algorithms in coding that identifies relationships in data, imitating a natural neural network — a brain. A graph neural network is a type of artificial neural network that uses graph data or a set of objects and their connections.

Fiona used coding algorithms to analyze the data from the US Patent Office, combining algorithms to form the visualization of her project provided in the graphic above. Essentially, Fiona created a visual representation of the data’s relationships. The US Patent Office data is a record of patents from 1975 to 2021. Because patents are essentially records of new inventions, Fiona could analyze patterns of innovation from this data. Fiona’s project grouped patents — really, inventions — by their structure and the inventor’s profile.

“This project was focused on several concepts; the first one is the innovation network,” Fiona said. “I used a graph neural network to explore the structure of innovation that you see in patterns and their citation relationships to characterize that link between innovations.”

Fiona focused her project on the Innovation Network article from a scholarly journal that was published by three Nobel Prize economists. Her interest in exploring innovation patterns developed in the summer before her junior year.

“I wanted to do this project because the summer before I started, I was in an entrepreneurship summer program where we got to learn about company protocols, and then we learned about the idea of patterns,” Fiona said. “I feel like innovation really is a part of my life and a part of what I do every day. I really wanted to find a way to characterize the innovation that’s currently happening in the world. When I thought about innovations and patterns, I realized there is a key connection between them, and that this connection can be characterized in some form, quantifying this abstract idea of innovation.”

Fiona attended the Harvey Mudd Computer Science program in her junior and senior year, where she met Dr. Qiwei Han, an assistant professor from the Nova School of Business and Economics, who helped her with her project.

“I think [working at Harvey Mudd] really helped me create a solid foundation in computer science and also form a bond with the professor, which also helped me with my project,” Fiona said.

Fiona started and completed her project during the most stressful time — her junior year during the pandemic. Coping with the impacts of the pandemic, schoolwork, and college applications, she struggled but eventually thrived while working on the project. She learned that scientific projects need repetitive work that might seem boring at first, but she would eventually feel accomplished after she finished everything. Scientifically, the project finds that the innovation network in this case is a validation spillover effect postulated by three economists, which gives a way for future public initiatives.

Her success also comes from the support she received from her mentor Dr. Han, who helped her complete the technical aspect of the project and dig deeply into the ideas. Fiona still vividly remembers the morning she found out she was named a Regeneron Scholar.

“It was a free block morning; I woke up two minutes before my alarm,” Fiona said. “I turned on my phone, and I got an email saying that I got named as a scholar. I was so shocked because I didn’t expect to get it. I was like ‘Oh my god! I got it.’”

After being awarded the title of Scholar, she has continued to pursue her passion in working on similar network analysis and combining computer science with social science in her future projects. She has the determination and confidence that her future will be bright by using the experience and knowledge she gained in this project.