題目：Big Data Research Opportunities in Disaster Management:
The LITMUS Landslide Information Service as an Illustrative Example
講師：Prof. Calton PuGeorgia Institute of Technology (Professor and J.P. Imlay, Jr. Chair in Software)
The ongoing convergence of evolution of devices (Internet of Things), deployment of large shared infrastructures (computing clouds), and accumulation of Big Data (sensors and social media) has created exciting new research challenges in quality of service (e.g., real-time response time, high availability and robustness despite widespread failures), and quality of information (e.g., security, privacy, and robustness despite misinformation). These research challenges require integration and synthesis of results from several related areas, e.g., sensor networks and social networks as Big Data producers, sophisticated machine learning models running on clouds as Big Data consumers, and system support at various levels to provide timely response. We have built LITMUS, a landslide detection system, to illustrate the research challenges of integrating physical sensors (earthquakes and rainfall) and social sensors (Twitter, Instagram, and YouTube). LITMUS is running live in the project website at [<grait-dm.org>].