about the lab
Research in the concept mining lab investigates how machine learning and analyses of language can provide insights into human thinking and mental health. We pursue research examining how Natural Language Processing and Big Data analyses can, in effect, data mine the mind.
In a paper recently published in PNAS and funded by the The John Templeton Foundation we investigated the impact of future thinking on decisions using big data analytics.
In a paper recently appearing in npj Schizophrenia -- Nature and funded in part by a Google Research Award we report how machine learning methods can be used to discover features of language that can predict the onset of a psychotic break as much as 2 years in advance. Both papers received significant Press coverage.
In paper to appear in Behavioral Research Methods, we use big data and machine learning to identify the emergence of mental illness from people's social media posts.
Currently, we are using deep learning models (e.g. BERT, Convolutional Neural Networks) to better understand decision making, the relationship between semantics and syntax, and the emergence of brain illness, such as alzheimer's,
Additional areas of research funded by NSF include the perceptual conditions that give rise to causal illusions and the role of forces in causal reasoning. The lab also pursues research in lexical semantics.