As crises such as natural disasters and pandemics start and the situation develops over time, a vast amount of news is generated through the recording of the social situation. The accumulation of such news data as a corpus and the construction of an archive with the analytics of the news corpus will serve as a basic infrastructure for utilizing the accomplishment of journalism for the study of society and its history. In this study, I will summarize three crisis news archives, typhoon, Great East Japan Earthquake, and COVID-19, and compare various methods for "reading" such a large-scale text corpus. We will also present some preliminary results on the characteristics of each crisis and comparisons among crises.

      @InProceedings{ daj21,
      author = {Asanobu KITAMOTO},
      title = {How to Read the Archives of Crisis News? - Comparison among Typhoons, 2011 Great East Japan Earthquake, and the COVID-19 Pandemic},
      booktitle = {Sixth Convention of Japan Society for Digital Archive},
      volume = {5},
      number = {s1},
      pages = {s51-s54},
      year = 2021,
      month = 4,
      doi = {10.24506/jsda.5.s1_s51},
      note = { (in Japanese)},
      }