Table of Contents
Fetching ...

Deflating the Chinese Balloon: Types of Twitter Bots in US-China balloon incident

Lynnette Hui Xian Ng, Kathleen M. Carley

TL;DR

This study tackles how Twitter bots participate in digital diplomacy during a high-tidelity cross-border incident by developing and applying an integrated pipeline to identify three bot types—General Bots, News Bots, and Bridging Bots. The approach combines BotHunter-based detection, content- and profile-based classification, Louvain clustering for bridging, and multi-faceted analysis (social network analysis, topic modeling, and the BEND information-manipulation framework) to quantify influence, themes, and maneuvers across US, China, and Rest of World. Key findings show that all three bot types are present in each geography, with US geotagged bots more focused on balloon-location narratives and Chinese geotagged bots on escalation; Chinese General and News Bots also exhibit higher centrality, while Bridging Bots are rare but pivotal for cross-border information flow. The work contributes transferable methods for analyzing digital diplomacy and bot-driven narratives, offering insights into how automated actors may shape public perception and policy discussions in international conflicts.

Abstract

As digitalization increases, countries employ digital diplomacy, harnessing digital resources to project their desired image. Digital diplomacy also encompasses the interactivity of digital platforms, providing a trove of public opinion that diplomatic agents can collect. Social media bots actively participate in political events through influencing political communication and purporting coordinated narratives to influence human behavior. This article provides a methodology towards identifying three types of bots: General Bots, News Bots and Bridging Bots, then further identify these classes of bots on Twitter during a diplomatic incident involving the United States and China. Using a series of computational methods, this article examines the impact of bots on the topics disseminated, the influence and the use of information maneuvers of bots within the social communication network. Among others, our results observe that all three types of bots are present across the two countries; bots geotagged to the US are generally concerned with the balloon location while those geotagged to China discussed topics related to escalating tensions; and perform different extent of positive narrative and network information maneuvers.

Deflating the Chinese Balloon: Types of Twitter Bots in US-China balloon incident

TL;DR

This study tackles how Twitter bots participate in digital diplomacy during a high-tidelity cross-border incident by developing and applying an integrated pipeline to identify three bot types—General Bots, News Bots, and Bridging Bots. The approach combines BotHunter-based detection, content- and profile-based classification, Louvain clustering for bridging, and multi-faceted analysis (social network analysis, topic modeling, and the BEND information-manipulation framework) to quantify influence, themes, and maneuvers across US, China, and Rest of World. Key findings show that all three bot types are present in each geography, with US geotagged bots more focused on balloon-location narratives and Chinese geotagged bots on escalation; Chinese General and News Bots also exhibit higher centrality, while Bridging Bots are rare but pivotal for cross-border information flow. The work contributes transferable methods for analyzing digital diplomacy and bot-driven narratives, offering insights into how automated actors may shape public perception and policy discussions in international conflicts.

Abstract

As digitalization increases, countries employ digital diplomacy, harnessing digital resources to project their desired image. Digital diplomacy also encompasses the interactivity of digital platforms, providing a trove of public opinion that diplomatic agents can collect. Social media bots actively participate in political events through influencing political communication and purporting coordinated narratives to influence human behavior. This article provides a methodology towards identifying three types of bots: General Bots, News Bots and Bridging Bots, then further identify these classes of bots on Twitter during a diplomatic incident involving the United States and China. Using a series of computational methods, this article examines the impact of bots on the topics disseminated, the influence and the use of information maneuvers of bots within the social communication network. Among others, our results observe that all three types of bots are present across the two countries; bots geotagged to the US are generally concerned with the balloon location while those geotagged to China discussed topics related to escalating tensions; and perform different extent of positive narrative and network information maneuvers.
Paper Structure (24 sections, 8 figures, 2 tables)

This paper contains 24 sections, 8 figures, 2 tables.

Figures (8)

  • Figure 1: Overview of Methodology for Identifying Types of Bots
  • Figure 2: Methodology for identification of News Bots
  • Figure 3: Methodology for identification of Bridging Bots
  • Figure 4: Proportion of User Types
  • Figure 5: User Type Distribution through Social Network Analysis
  • ...and 3 more figures