自然灾害、社会安全、公共卫生等大规模突发事件已经成为当前中国的主要社会问题之一。然而，由于数据的局限性，我们对于大规模突发事件下受影响人群的行为模式缺乏全面的了解和科学的认识。本项目旨在依托申请者以往的研究基础和数据渠道，通过结合手机数据和在线社会媒体数据，对大规模突发事件情况下的群体行为动力学进行深入分析，重点研究基于地理位置信息的移动行为模式和基于相互交流信息的通信网络与在线社会网络的结构和演化特征；在此基础上，一方面，将信息学、控制学等领域的信号监测模型引入项目中，对人群移动和网络活动的异常行为进行监测，另一方面，建立人群移动和交流行为的预测模型；最后，综合两种数据源，对得到的模型进行相互验证，并分析手机数据模型与在线社会媒体数据模型的相关性。本项目有望为我国应急管理理论与方法增添丰富的一笔，并促进大数据在我国减灾救灾、处置突发公共安全事故等大规模突发事件中的应用。

Large-scale emergencies, such as nature disasters, social security and public health, have become one of the most challenging social problems in China. Due to the limitation of data sources, it is not well understood for the behaviors of affected people. With facilitated data access developed by previous research, this project aims to investigate human dynamics under conditions of large-scale emergencies, by integrating mobile communication and online social media data sources. We focus on the analysis of geo-location based human traveling behavior, as well as the structure and dynamics of communication networks and online social networks. Based on this, on the one hand, we introduce signal monitoring models from disciplines of information and control theory, to detect abnomities of human traveling behavior and network activities. On the other hand, we build models for the prediction of human behaviors under study. Lastly, we evaluate model results from one data source by veryfying it with the other datasource, and we analyze the correlation between mobile phone data based models and online social media data based models. This project is expected to contribute critical theory and methodology for emergency management in China, and it has the potential to promote the application of Big Data in large-scale emergency management, such as in disaster mitigation and relief, and in pubic safty accident management.