HCC: Small: The Evolution of Language Use in Online Communities

University Of Southern California, Los Angeles CA

Investigators

Abstract

This research seeks to understand the lifecycle of hate language in online communities: how people's use of hateful language evolves over time and across groups, and whether or when members leave such groups. By analyzing large datasets from various social media platforms using advanced artificial intelligence (AI) tools, the project team will identify the key factors that influence the evolution of participation in groups that tolerate hateful language. This research will bridge social science and AI, offering new insights into online polarization. The team will also develop tools that others can use to replicate and extend the analysis, and that can also be used in interdisciplinary courses and summer schools aimed at training future researchers in this area. This project includes developing predictive models to detect initial involvement in a hate language group that leads to participation in others, and evaluating the impact of inter- and intra-group conflicts on user behavior. The analytical approach involves using explainable AI techniques based on Large Language Models to interpret model predictions and causal inference methods to understand the influence of external events on the consumption and production of hateful language in online groups. The developed tools will enable the research team -- and other researchers -- to analyze the language and behavior of users within groups that tolerate hateful language, identify key features predicting membership in these groups, and understand the causal mechanisms behind those processes. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.