论文标题

通过aweb进行协调:分析来自中国,伊朗,俄罗斯和委内瑞拉的基于图像的影响力运营

Coordinated through aWeb of Images: Analysis of Image-based Influence Operations from China, Iran, Russia, and Venezuela

论文作者

Ng, Lynnette Hui Xian, Moffitt, J. D., Carley, Kathleen M.

论文摘要

国家赞助的在线影响力通常包括协调的帐户,利用在线空间影响公众舆论。与这些操作相关的帐户使用图像和模因作为其内容产生的一部分和传播策略的一部分,以提高内容的有效性和参与度。在本文中,我们介绍了来自源自中国,伊朗,俄罗斯和委内瑞拉国家的Phomemes 2022挑战的图像的研究。首先,我们通过量化图像相似性来分析每个国家内部和整个国家内部图像的协调。然后,我们构建图像图像网络和图像簇,以识别图像中的关键主题影响操作。我们得出相应的帐户账户网络,以可视化每个国家 /地区参与帐户之间的交互。最后,我们在每个国家的组织运营的更广泛背景下解释了图像内容和网络结构。

State-sponsored online influence operations typically consist of coordinated accounts exploiting the online space to influence public opinion. Accounts associated with these operations use images and memes as part of their content generation and dissemination strategy to increase the effectiveness and engagement of the content. In this paper, we present a study of images from the PhoMemes 2022 Challenge originating from the countries China, Iran, Russia, and Venezuela. First, we analyze the coordination of images within and across each country by quantifying image similarity. Then, we construct Image-Image networks and image clusters to identify key themes in the image influence operations. We derive the corresponding Account-Account networks to visualize the interaction between participating accounts within each country. Finally, we interpret the image content and network structure in the broader context of the organization and structure of influence operations in each country.

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源