I am a fifth year graduate student in the Psychology Department at the Uiniversity of Illinois Urbana-Champaign, and a member of Learning & Language Lab, advised by Dr. Jon Willits.
Before diving into the world of psychology, I had eight years of experience in mathematics: After receiving a B.S. degree in math, I continued on studying theory and application of random graph as a Ph.D student at University of Illinois Urbana-Champaign, where I had exposures to diverse math topics, e.g. Probability theory, Mathematical Logic, Function Analysis and basic Algebra. At the meantime, I was profoundly interested in formal linguistics and received trainings in syntax and formal semantics. My education in mathematics and linguistics has benefitted me with a broad background and relatively deep understanding in formal systems, alongside with intuitions to modeling cogntivie stuctures and processes using formal tools.
Back to psychology, my research interests include semantic memory, representation of langauge meaning (for both words and complex structures like phrases and sentences), and I study the topics with theoretical and computational modeling. In particular, I have been working on developing a semantic memory model which encodes distributional data from raw linguistic input, in structured graphical forms, e.g. constituent trees, with the goal to represent the meaning of phrases and sentences generated from the productive natural language.
Publications
Mao, S., Huebner, P., & Willits, J. A. (2023). Spatial versus graphical representation of Distributional semantic knowledge. Psychological Review, in press. pdf
Mao, S., Huebner, P. A., Willits, J. A. (2022). Compositional generalization in a graph-based model of distributional semantics. In Proceedings of the 44th Annual Conference of the Cognitive Science Society. (pp. 1993-1999). pdf
Under Review
Mao, S., Huebner, P. A., Willits, J. A. (under review). Conditions Underlying Success and Failure of Compositional Generalization in Distributional Models of Language.
Mao, S., Huebner, P. A., Willits, J. A. (under review). Structured Graphs Built from Distributional Information Facilitate Compositional Generalization in Language.