Meaningful Open Source Indicators (MOSI)
Led by quantitative social scientist Catie Snow Bailard, this cluster focuses on analyzing and developing computational tools for identifying and tracking harmful content online and the malicious actors who produce such content. Meaningful Open Source Indicators (MOSI) combines the subject-area and technical expertise of researchers working across a wide range of disciplines—including engineers, data scientists, computer scientists, computational social scientists and traditional social scientists. This interdisciplinary approach is an essential component of MOSI’s goal to conduct computational analyses that capture the richness and complexity of human behavior, interactions and social dynamics online; rather than merely capturing what’s easiest to detect and measure automatically.
MOSI researchers are also united in their focus on investigating the use of digital technologies to disseminate, promote and orchestrate deleterious online content and behavior; particularly that which has the potential to spill over into offline harms. Under this umbrella, our current focal point is the dissemination of m/disinformation and abusive content, such as hate speech, through online platforms. In addition to conducting original research and developing novel tools to this end, MOSI researchers also endeavor to test the validity and reliability of existing tools designed to track and monitor malicious content and behavior online. This includes explorations of how they are used by practitioners, the explainability of tool output and core definitions of online harms. The ultimate objective motivating research conducted by the MOSI cluster is to generate empirically-grounded, actionable insights and technological approaches to aid researchers, journalists, policymakers and technologists in the abiding effort to monitor and counter digital harms and the threat they pose to democratic societies