There are four main research themes in the ACTION lab.
1. Human Cognition with Evolving Technologies
How do people across the lifespan learn new knowledge outside of the classroom? How do people across the lifespan search and assess the “evidence” needed to make complex judgments?
The research program aims at advancing basic knowledge in cognitive sciences about how human minds interact with evolving technologies (including recommender systems, AI, etc.) through self-regulated processes, such as information search and learning, for complex information goals, including decision-making and knowledge acquisition. Current projects include: search as learning/ Information foraging across the lifespan; exploring knowledge representations with large language models.
2. Coupling Cognitive Agents
How do people + machine work together to perform complex cognitive tasks? How can we build a system to help both machine and human learn new knowledge?
The research program aims at translating theories in cognitive psychology, human-computer interaction, machine learning, and language models to design intelligent agents (including multi-agent environments) that support individual and group human behavior (such as learning, decision-making, health behavior, ideation creation) for adults across the lifespan. Current projects include: developing cognitive generative agents with theory-driven memory and reasoning systems; developing intelligent agents and designing human-agent interaction to support ideation and creativity; designing and developing pedagogical agents to support self-regulated learning.
3. Understandable and Actionable Information for All
How can we refine large language models to deliver understandable and actionable health information for older adults with inadequate health literacy?
How can we use AI to deliver motivational messages to promote behavior change?
The research program aims at building socio-technical solutions to promote the implementation and dissemination of understandable, actionable, and persuasive information for all. Current projects include: improving large language models (LLMs) to generate motivational, persuasive, and actionable information; refining large language models to improve comprehension and promote health literacy of older adults.
4. Digital Health for All
How can we use innovative technologies to help people with chronic and progressive illness to stay healthy on their own?
This research theme is to leverage behavioral sciences and data sciences to establish novel generalizable user models of information behavior to support the development of personalized digital health innovation for promoting health behavior (such as self-management or behavior change) in the naturalistic environment and addressing health disparities. Current projects include: developing motivational interviewing conversational agents to support health behavior change and depression regulation; building e-Lifestyle coaches to augment self-regulated health promotion with evolving needs.