Emergency Operation Center (EOC) must have a Common Operational Picture (COP) to respond efficiently to disasters and crises. Not only national / local government officers in EOC but also disaster researchers need COP about disaster summary and current situation. There are enormous amount of information on disaster so that a lot of time and effort must be put for processing it. We have developed an information processing system which automatically extracts keywords from the digital text dataset acquired from web-news on disasters. This system aims at visualizing a "big picture" for decision making by emergency responders and disaster research by researchers. In this paper, we report a web publication service project using automatic keyword extraction and visualization system. This service have been used heavily by emergency responders, managers, and researchers since this service have added an automatic selection function of recommended news articles in each operational period and push-type information delivery system using e-mail letter which reports their titles and keywords set.

View full abstract