Collaborative Research: HNDS-R: Polarization, Information Integrity, and Diffusion

这个项目的目的是加深对两极分化的理解，两极分化是误导言论传播的漩涡。社交媒体对公民的承诺是，人们可以普遍获得必要的知识，从而为有争议的问题提供信息并激发广泛的政治参与。相反，社交媒体已经成为认知两极分化的培养皿——将公众划分为红色和蓝色的现实。两极分化的、无法核实的言论在社交媒体上的传播，侵蚀了建设性分歧所需的共享知识，并让外国对手得以发起具有可怕病毒式传播能力的有害传播活动。通过对新旧社交媒体数据的分析和新的分析方法的开发，本项目的研究使人们对误导性言论的传播动态有了更深入的了解。这将有助于提高网络空间的可信度。利用之前收集的Twitter十年档案，以及即将收集、整理和匿名化的新数据，本项目调查了两极分化对在线网络中误导性陈述传播的影响。新的数据集将是同类中最大的公开可用数据集，有助于未来对这一主题的研究。此外，还开发了分析方法来评估意识形态内容、信息的准确性、用户的一致性以及误导性陈述级联的结构和时间特征，如覆盖范围、速度和病毒式传播。这项工作的结果是对误导性陈述的扩散动力学有了更深入的理解。这可以为早期遏制对策提供信息，以便在传染病蔓延失控之前更有效地识别、跟踪和消除扩散。本项目由社会学项目、人际网络与数据科学研究（HNDS-R）项目和安全与可信网络空间（SaTC）项目共同资助。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The aim of this project is to deepen understanding of polarization as the vortex in which misleading statements go viral. The civic promise of social media was universal access to the knowledge needed to inform and motivate broad political engagement on controversial issues. Instead, social media have become the Petri dish for epistemic polarization – the division of the public into red and blue realities. The spread of polarizing unverifiable statements over social media erodes the shared knowledge needed for constructive disagreement and allows foreign adversaries to mount toxic communication campaigns with frightening virality. Through analysis of old and new social media data and development of new analytic methods, the research in this project brings a greater understanding of the diffusion dynamics of misleading statements. This will be useful in improving the trustworthiness of cyberspace.Using previously collected ten-year Twitter archive, along with new data to be collected, curated, and anonymized, in this project the impact of polarization on the dissemination of misleading statements in online networks is investigated. The new data set will be the largest publicly available data set of its kind, contributing to future research on this topic. In addition, analytical methods are developed to evaluate the ideological content, veracity of messages, alignment of users, and the structural and temporal characteristics of misleading statement cascades, such as reach, velocity, and virality. What results from this work is a deeper understanding of the diffusion dynamics of misleading statements. This can inform an early-containment counterstrategy that can more effectively identify, track, and neutralize diffusion before the contagion can spread out of control. This project is funded jointly by the Sociology Program, Human Networks and Data Science – Research (HNDS-R) Program, and Secure and Trustworthy Cyberspace (SaTC) Program.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.