about the lab
Research in the Concept Mining Lab centers on the conceptual system and its influence on everyday reasoning.
In a line of research funded by Google Research, we have analyzed several large bodies of text (e.g., Wikipedia) to uncover patterns of meaning across thousands of words in English. These patterns are then compared against the judgments of hundreds of participants to establish their psychological validity and to see what they might indicate about the large-scale structure of the conceptual system.
In a line of research funded by The John Templeton Foundation and the University of Pennsylvania, we have been using big data techniques to examine people’s thoughts about the future and how these thoughts might be related to wellbeing. By analyzing millions of tweets, we have been able to determine the degree to which different states in the US talk about the future as well as show that these different levels of future orientation are associated with different levels of physical and mental health. In conjunction with this research, we have also adapted neural network approaches from machine learning to identify different categories of future oriented thought and then show how these categories of thought are associated with personality and characteristics of well-being.
In a third line of research funded by NSF, we have examined the characteristics of everyday reasoning by examining the perceptual conditions that give rise to causal illusions. Causal illusions stand at the core of pseudoscientific beliefs and quackery and reveal themselves in the harmless rituals and superstitious behaviors people perform before sporting events or tests. Among other phenomena, we have examined the time course of mental processes involved in seeing a person “magically” open an elevator door. In a related project, we have examined how “seeing” forces can affect people’s sensitivity to actual physical forces generated by a haptic controller device (effectively, a robotic arm).
Research in the CLS lab draws on several different methodologies, including computer visualization, cross-linguistic comparisons, large-scale corpus analyses, and haptic rendering. While much of our research is tied to language, there are also a number of research projects--in particular, those focusing on causation, space, and intention--that only concern non-linguistic conceptual and perceptual phenomenon. Much of our research involves the development and testing of computational models, as well as, the study of special populations, such as bilinguals. We draw on insights from psychology, linguistics, philosophy, computer science, and cognitive neuroscience.