Catalyzing Collaborations: A Model for the Dynamics of Team Formation at Conferences

The COVID-19 pandemic has brought to the fore the importance of collaboration among scientists to address global challenges. One of the main ways that collaborations are catalyzed is by gathering scientists together at conferences. In the U.S. alone, conferences amount to billions of dollars per year in terms of travel expenses, organizing costs, and loss of research time. In this work, I present a dynamical model for predicting the formation of scientific collaborations at conferences, inspired by the process of catalysis. Specifically, the model tracks the probability that conference participants form collaborations given their level of interaction throughout the conference. Model predictions are tested using data from several multi-year series of interactive conferences known as Scialog Conferences, organized by the Research Corporation for Science Advancement over the period 2015-2020. We find that scientists who interact more intensely throughout the conference have a higher likelihood of forming a collaboration. Furthermore, we find that the likelihood of collaborating remains at a higher level even after the interaction between participants has ceased.