RIDIR: Collaborative Research: Integrated Communication Database and Computational Tools

该项目将为社会技术网络安全研究和行为、信息和计算机科学家对信息来源的更广泛调查开发一个综合研究框架。目前，研究人员主要局限于对大量在线文本的自然语言处理。该项目将使分析更大的信息世界成为可能，包括来自中国等国家的信息，以及报纸、电视和在线资源中的信息流动，包括视频和音频信息。该项目解决了网络安全研究的一个核心目标，即了解信息战和审查的来源、流动和终止。该项目旨在建立一个集成和统一的信息数据库，该数据库将来自六个地点的电视和印刷来源的大众传播数据与两个流行的在线传播平台的数据相结合。该项目将通过使用文本、图像、视频和音频的自动化多模式内容分析，生成各种关于主题、演员、事件和情感的元数据和时间序列数据。将变量联系起来，以确定通过多个平台的通信渠道之间的信息流动轨迹。它将开发一类新的计算模型和算法，可以通过机器学习、计算机视觉、深度学习和自然语言处理自动分析口头和非口头通信数据。该项目将允许计算科学和社会科学的研究人员通过定性研究的搜索界面、定量研究的统计包和各种可视化工具访问元数据和时间序列数据。因此，该项目将使用尖端的计算方法连接以前未开发的数据源，使学者能够对新兴信息和通信生态系统中的大规模模式进行系统研究。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。

This project will develop an integrated research framework for sociotechnical cybersecurity research and broader investigations of information provenance by behavioral, information, and computer scientists. Currently, researchers are mainly limited to natural language processing of large bodies of online text. This project will make it possible to analyze larger information worlds, including those from such countries as China, and the flow of information, including video and audio information, in newspapers, TV, and online sources. The project addresses a core goal of cybersecurity research, which is to understand the provenance, flow, and termination of information warfare, and censorship.The project is aimed at constructing an integrated and unified information database that combines mass communication data from TV and print sources from six locations, with data from two popular online communication platforms. The project will generate a variety of metadata and time series data on topics, actors, events, and sentiments presented in communications by automated multimodal content analysis using text, image, video, and audio. Variables will be linked to identify trajectories of information flow between communication channels through multiple platforms. It will develop a new class of computational models and algorithms that can automatically analyze both verbal and nonverbal communications data by machine learning, computer vision, deep learning, and natural language processing. This project will allow researchers across the computational and social sciences to access the metadata and time series data through a search interface for qualitative research, a statistical package for quantitative research, and various visualization tools. This project will therefore link previously untapped data sources using cutting-edge computational methods to enable scholars to conduct systematic research on large-scale patterns in the emerging information and communication ecosystem.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.