IBSS: Spatiotemporal Modeling of Human Dynamics Across Social Media and Social Networks

这一跨学科研究项目将研究社交媒体和社交网络中的人类动态，并将重点放在信息在时间和空间上传播的建模上。它还将研究在线活动与现实世界人类行为之间的联系。由于研究团队将采用广泛的不同研究方法来研究人们的行为在电子媒体中反映的过程，该项目将产生新的传播理论、知识发现工具和计算模型。它将促进不同领域的信息和见解的融合，包括空间科学、社交媒体、通信、计算机科学和社会行为分析。该项目还将促进在行为和社会科学研究中采用计算应用程序和建模方法。调查人员将对两种不同的情景(州和国家层面对灾害警告和警报的公众反应以及有争议的社会话题的全民公投)进行测试，以提炼出一种处理模因或可重复信息的传播新理论。该项目的一个产品将是与圣地亚哥市紧急服务办公室协调设计和测试的社交媒体推广平台原型。该原型将促进在灾害事件期间通过多种社交媒体渠道向目标人群快速传播官方警报和警告通知。该平台将根据社交网络影响因素确定和招募1000名社交媒体志愿者，从而使政府机构能够更有效地与公众沟通，并更好地为自然灾害和与人类有关的危机做好准备。该项目还将为多个学科的学生提供研究和教育机会，包括地理、语言学、计算机科学、社会科学和传播学。人们在时间和空间上移动，在这样做的同时，他们也会进行交流。有了现代通信技术，他们在日常行为中留下了痕迹，并引用了他们的行为。这些信息回声具有巨大的潜力，可以为人们提供对行为的洞察。研究小组将使用计算方法，如社会网络分析、地理信息系统和机器学习，以及传统的社会科学研究方法，如定性分析、推理统计和行为分析，分析人类消息、活动和通信的传播模式。研究人员将建立一个跨学科的研究框架，从时空建模的角度研究人类动力学和信息传播。他们将验证和改进Meme扩散(M3D)传播理论的多层次模型，以通过社交媒体和社交网络进行在线人类交流。他们将使用计算预测方法和基于代理的建模(ABM)方法，结合人类动力学的两个场景(灾难警告/警报和有争议的社会话题的公投/主张)来分析时空模式的动态变化。他们还将结合高性能计算、基于Web的地理信息系统工具、基于代理的建模和开源软件，为社会科学家开发有效和可访问的数据处理、可视化和分析工具，以研究人类动力学和信息传播。该项目通过NSF跨学科行为和社会科学研究(IBSS)竞赛得到支持。

This interdisciplinary research project will examine human dynamics across social media and social networks and will focus on the modeling of information diffusion over both time and space. It also will study the connection between online activities and real world human behaviors. Because the research team will employ a broad range of different research approaches to study the processes through which people's behavior is reflected in electronic media, the project will yield new communication theories, knowledge-discovery tools, and computational models. It will facilitate the convergence of information and insights across a diverse set of fields, including spatial science, social media, communication, computer science, and social behavioral analysis. The project also will facilitate the adoption of computational applications and modeling approaches in behavioral and social science research. The investigators will conduct tests of two different scenarios (public response to disaster warnings and alerts and referendums of controversial social topics at state and national levels) to refine a new theory about communication dealing with memes, or reproducible messages. One product of the project will be a prototype social media outreach platform designed and tested in coordination with the city of San Diego's Office of Emergency Services. The prototype will facilitate rapid dissemination of official alerts and warnings notifications during disaster events via multiple social media channels to targeted populations. The platform will be designed to identify and recruit 1,000 social media volunteers based on their social network influence factors, thereby enabling government agencies to communicate more effectively to the public and to be better prepared for both natural disasters and human-related crises. The project also will provide research and educational opportunities for students in multiple disciplines, including geography, linguistics, computer science, social science, and communication.People move in time and space, and while they do so, they communicate. With modern communication technologies, they leave traces of and refer to their behavior while they engage in everyday behavior. These informational echoes have enormous potential to provide insights into people's behavior. The research team will analyze the diffusion patterns of human messages, activities, and communications using computational methods, such as social network analysis, geographic information systems, and machine-learning, as well as traditional social scientific research methods, such as qualitative analysis, inferential statistics, and behavior analysis. The investigators will build an interdisciplinary research framework for studying human dynamics and information diffusion from a spatiotemporal modeling perspective. They will validate and improve the Multilevel Model of Meme Diffusion (M3D) communication theory for online human communications across social media and social networks. They will analyze the dynamic changes of spatiotemporal patterns with two scenarios of human dynamics (disaster warnings/alerts and referendum/propositions of controversial social topics) using computational predictive methods and agent-based modeling (ABM) approaches. They also will develop effective and accessible data processing, visualization, and analytical tools for social scientists to study human dynamics and information diffusion by combining high-performance computing, Web-based geographic information system tools, agent-based modeling, and open-source software. This project is supported through the NSF Interdisciplinary Behavioral and Social Sciences Research (IBSS) competition.