Sandia National Labs Academic Programs Collaboration Report

misinformation/disinformation and influence operations. Both the physics-inspired work and the COVID-related work were presented at the Tenth International Conference on Complex Systems. Collaborations with Professors Yuguo Chen at Illinois and Abdullah Mueen at the University of New Mexico (UNM) were significant to the project. Illinois developed a nonparametric clustering model for dynamic networks with an evolving group structure, extending existing latent space capabilities by explicitly modeling the birth, death, splitting, and merging of groups with a hierarchical Dirichlet process hidden Markov model. The project’s UNM partner developed a new method for using tweet metadata to predict content. The results of this project will be important to NNSA and the DOD since it showed that multiscale group dynamics exist and developed an initial set of new theories and models of those dynamics. Now that multiscale group phenomena do in fact exist, a new line of research in this area can be undertaken to leverage knowledge of these dynamics for national security and other applications.

Yuguo Chen is a professor in the Illinois Department of Statistics and the Director of Illinois Statistics Office . He was instrumental in helping Sandia develop a nonparametric clustering model for dynamic networks with an evolving group structure for the LDRD project that utilized emergent recursive multiscale interaction project. Chen commented, “New tools for understanding the sequential evolution of communities in dynamic networks, such as complex social systems, can be built on our methodology.”

Group dynamics modeled using new communication vibration theory: (1) Movement of members between days; (2) changes in the average distance between group centroid and members between days; and (3) movement of centroid of group members relative to non-members.

61

2021-2022 Collaboration Report