A Signal-Processing Approach to Crowd Reactions in Political Rallies
Dr Alejandro Canek Hermida Carrillo (Imperial College London) explores how acoustic signals can help us advance the study of crowd dynamics in political events.
Abstract
All around the world, crowds mobilise to show support for ideological leaders and causes. These events generate rich collective dynamics—chants, applause, booing—that are often interpreted as indicators of ideological alignment or dissent. Yet the empirical tools commonly used by social scientists to analyse such crowd reactions remain limited, frequently relying on third-party transcripts, manual coding, or qualitative interpretation.
This presentation introduces advances in the development of a signal-processing–based methodology for analysing crowd dynamics in political events. The approach separates crowd-related acoustic activity from other sources, enabling the extraction of quantitative features that capture the timing and intensity of crowd reactions.
Dr Hermida Carrillo outlines the methodological framework and presents an initial assessment of what acoustic features might reveal beyond the current state of the art in social-scientific research. He concludes by highlighting current challenges and exemplary research questions that this approach enables.
This project is conducted in cooperation with the Machine Listening Lab at Queen Mary University of London.
About the speaker
Alejandro Canek Hermida Carrillo is an Imperial College Research Fellow (ICRF) in the Department of Management and Entrepreneurship at Imperial Business School. He is also an affiliated researcher with the Computational Culture Lab (Stanford/Berkeley)
In his fellowship, he combines computational methods with sociological and social psychological theories to examine the co-evolution between crowds and the rhetoric of autocratic leaders.
Alejandro holds a PhD in Management from the LMU Munich, an MSc in Economic, Organisational, and Social Psychology also from the LMU Munich, and a Licenciatura in Psychology from the National Autonomous University of Mexico (UNAM).
This event is part of the Financial Computing and Analytics Research Group seminar series at UCL Computer Science.
