The modern information environment, defined by its physical, informational, and cognitive dimensions, is increasingly dominated by dynamic and unpredictable as- pects of the internet. The internet enables the free flow of information, yet can be strategically manipulated by different types of actors deploying complex information campaigns through social networks and news mediums to persuade and change the way that people think. The vast influence and stealthy nature of these campaigns, especially those orchestrated by China, are critically understudied due to proprietary data restrictions, resource constraints, and the anonymity of digital interactions, leaving a significant research void in identifying and analyzing state-sponsored inauthentic behaviors on social media platforms.

Today's internet plays a crucial role in the information environment, where information can be disseminated widely through viral tweets, controlled via state censor protocols, or suppressed through government internet censorship. State actors are leveraging social networks for information campaigns, making it imperative to understand who is disseminating what to whom. However, existing research on state-sponsored campaigns, particularly related to China's activities, faces challenges due to proprietary data restrictions, resource constraints, and the anonymous nature of many public forums, resulting in a gap in identifying state-sponsored inauthentic behavior on social media.

To address this research gap, this thesis leverages computational, network sci- ence , and natural language processing methods with political and sociological groundings to identify and characterize state actor campaigns within the information environment. At the singular campaign level, I use the network structure of a campaign to identify key actors and their functions across different types of information campaigns. I then employ information theory methods and algorithmic analysis to study covert elements for information campaigns, using a benchmark dataset to ground this section that is by its nature difficult to attribute to any actor. The use of narratives within the information environment is as critical to understand as the types of network impacts they are meant to create on a targeted audience. I analyze how China's dialogue towards regional countries, friends and foes have shifted, specifically before and after Russia's invasion into Ukraine, examining the aggregated effects of campaign narratives. Furthermore, I expand the research by utilizing a social cybersecurity framework to map out how China uses information maneuvers to shape the information environment. Lastly, I aggregate my findings by expanding a decision-making game designed for analysts who are studying how different types of actors manipulate the social media domain. I integrate key measurements in the synthetic creation of a PRC information campaign and validate generated data. These contributions enhance the understanding of state-sponsored information campaigns provide valuable tools to researchers in the field.

pages

Thesis Committee:
Kathleen M. Carley(Chair)
Patrick Park
John Chin
Colonel Matthew Benigni (U.S. Army)

Nicolas Christin, Head, Software and Societal Systems Department
Martial Hebert, Dean, School of Computer Science

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu