RAPID: Early Detection of Disease Outbreaks using Self-Organizing Patterns – COVID-19

该项目将通过开发一个强大的、可预测的模型来推动国民健康，该模型通过创新地整合公共卫生信息和社交媒体数据来图形化地表示新冠肺炎的传播。该项目开发了一个预测性建模工具，利用人工智能技术支持数据收集、分析和结果表示，以可视化地表示新冠肺炎或其他潜在流行病的传播。如果研究人员成功地开发出一种模型，利用传统的公共卫生数据与社交媒体数据相结合，迅速创建对疾病传播的可靠预测，那么社会效益将是显著的。该项目将有助于建立一个可在全球任何地方访问的开放源码数据库，同时向政府机构提供预警、检测和信号。这个快速项目开发了一个大规模的流行病模型，以实现数据共享，使用基于人工智能的方法和预测性建模。具体地说，该项目提出了一个快速和早期检测疾病的模型，将三种新的智能方法结合在一起进行暴发检测。这三种方法包括1)自组织系统理论，以检测新的模式形成；2)在研究交流中利用主题邻近，而不是受感染个体的地理空间邻近；以及3)主题网络中复杂性的丧失，作为即将爆发的“不健康”系统的指示器。在追踪疾病暴发的努力中，以前从未尝试过这种综合方法。拟议的方法预计将在项目的第二季度内产生一个工作模型。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势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 advance national health through the development of a robust, predictive model to graphically represent the spread of COVID-19 through an innovative integration of public health information and social media data. This project develops a predictive modeling tool to visually represent the spread of COVID-19 or other potential pandemics utilizing artificial intelligence techniques to support data gathering, analysis and representation of the outcomes. The societal benefit is significant if the researchers are successful in developing a model that utilizes traditional public health data integrated with social media data to expeditiously create a reliable prediction of disease spread. This project will contribute towards building an open source database that can be accessed anywhere across the globe while providing early warning detection and signals to government agencies. This RAPID project develops a large-scale pandemic model to enable data sharing, using AI-based approaches, and predictive modeling. Specifically, the project proposes a model for rapid and early disease detection, by combining three novel intellectual approaches to outbreak detection. These three approaches include 1) Self-organizing systems theory to detect nascent pattern formation; 2) Leveraging topical proximity in research communications over geospatial proximity of infected individuals; and 3) Loss of complexity in topical networks as an indicator of an “unhealthy” system with an impending outbreak. This integrated approach has not been previously attempted in efforts to track disease outbreak. The proposed methodology is expected to produce a working model within the 2nd quarter of the project.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.