Abstract

How do states attempt to shape public opinion in the contemporary era? Digital technology enables sophisticated yet relatively cheap online influence campaigns (OICs) that spread disinformation and exploit societal divisions, especially in democracies. Current events indicate that OICs are hard-to-detect yet influential in both domestic and international politics. Our research examines the nature of these operations on Twitter, a popular social media platform where at least 15 states have operated bot accounts within the past decade. Specifically, we analyze the factors that explain variation in bot accounts’ activity and the tactics

that they use to influence Twitter users. Our approach consists of automated textual analysis techniques and developing a machine learning model to classify a new dataset that we have compiled from Twitter data of all approximately 13 million Tweets identified as originating from state-operated bot accounts. In our research, we will also contribute to the public good by developing a publicly accessible Shiny application that permits researchers and organizations to access the results of our analyses. Overall, our project will contribute to knowledge on the role of OICs in politics and enhance future efforts to detect them on Twitter.

Principal Investigators