Description of original award (Fiscal Year 2023, $953,203)
Goals: Investigating, reducing, and preventing mis-, dis-, and mal-information (MDM) has become a top priority for governments, private organizations, and researchers. Nationally publicized political events often become focal points of MDM, which are exploited by various individuals and groups to launch disinformation campaigns and trigger spontaneous or crowd-sourced diffusion of disinformation and violent extremism. In this context, this project seeks to answer the following overarching questions: (1) What characteristics of high-profile events are more likely to trigger online MDM campaigns? (2) What network patterns are associated with online MDM clusters, and dissemination pathways, and what are the technical characteristics of more vulnerable platforms? (3) What are the common characteristics of organizations and other actors engaged in MDM campaigns? (4) What are the key themes and linguistic features of MDM campaigns, and how do they vary across disseminators and adopters?
Research Design: The above questions will be answered by conducting the first real-time mapping of the spread of MDM campaigns around contentious public events. Accordingly, the proposed team will (a) develop specialized algorithms to identify the creation of MDM campaigns and capture event-level characteristics of real life events which trigger MDM, (b) identify online MDM clusters, and dissemination pathways, and (c) use computerized linguistic tools for MDM content and sentiments analysis, and examine its association with users’ behaviors. In the first phase, the team will produce an original events-level dataset that documents real-time characteristics of MDM-triggering political events. In the second phase, advanced computational tools will be employed to monitor MDM campaigns in near–real time and to identify specific accounts/nodes linked to outsized contributions. The team will also map MDM network patterns, identifying variation across creators/adopters, and complete computer-assisted linguistic and sentiments analysis of MDM content.
Impact: The outcomes of the proposed project will help shape responses to MDM, and will be useful for policy makers, law enforcement officials, as well as community stakeholders. To maximize impact, the team will develop outreach materials and an online dashboard with an MDM tracker that allows users to understand real-time campaigns and their diffusion. The dashboard will provide a multi-level framework to understand the determinants and characteristics of MDM outbreaks, and identify links between specific linguistic characteristics and their virality. Additionally, the MDM tracker will provide insights into the commonalities and differences amongst online behaviors of MDM adopters, non-adopters, and disseminators. CA/NCF