The potential of big data-based disaster reduction was clearly recognized after Great East Japan Earthquake in March 2011. We claim that the ultimate goal of big data-based disaster reduction is to create a disaster media which for any people relevant information for themselves is available for use at the right time. To create a media that takes advantage of big data, we have been focusing on natural disaster data such as meteorological and earthquake data, and studying on both individual research challenges such as collection, analysis, distribution, and acceptance of big data, and their integration. A realization of such a media may lead to a solution for a typical problem in recent disasters; namely “a person in charge was overwhelmed by too much information,” through automation by artificial intelligence. At the same time, however, the goal of disaster reduction is to “save human lives by urging actions to humans,” and big data-based disaster reduction without any consideration on human factors is meaningless. Hence the talk combines an overview of past research and its perspective within pattern recognition and other domains, with future perspective on the important challenge of dividing and sharing roles between humans and machines.

      @InProceedings{ k:prmu16,
      author = {Asanobu KITAMOTO},
      title = {Big Data-Based Disaster Reduction, and the Role of Humans and Machines},
      booktitle = {Technical Report of Institute of Electronics, Information, and Communication Engineers (IEICE)},
      volume = {115},
      number = {456},
      pages = {131-131},
      year = 2016,
      month = 2,
      doi = {},
      note = { (in Japanese)},
      }