[M-IS22] History X Earth and Planetary Science

[MIS22-P05] Geographical Information in the Database of Materials for the History of Japanese Earthquakes

*Yasuyuki Kano^1,2, Junzo Ohmura^1,2 (1.Earthquake Research Institute, The University of Tokyo, 2.Collaborative Research Organization for Historical Materials on Earthquakes and Volcanoes, The University of Tokyo)

Keywords:Historical earthquakes, Historical materials on earthquake, Database, Place name, Historical Administrative Region Dataset β

Database of materials for the history of Japanese earthquakes is published by Collaborative Research Organization for Historical Materials on Earthquakes and Volcanoes, The University of Tokyo. The database provides a keyword search function in the text and source literature for materials related to earthquakes and volcanic eruptions. Each section of the series of earthquake record has a summary at the beginning. The summary consists of date of the earthquake and eruption event, geographical area affected by the event. The summary can also be used for keyword search. Regional distribution of damage is important for analyses on earthquake and eruptions. It is desirable that geographical information be provided, such as which places are historical materials or which places are written in the historical materials. It is desirable that the database provide geographical information such as location of historical materials and places written in the historical materials. In this database, we are trying to add geographical information in two ways: use location and name of a place that (1) appears in source literature field, and (2) in the text. For (1), municipality name can be extracted from title of municipal history especially in the case of municipal histories. The municipality names are searched on “Historical Administrative Region Dataset β” and is associated with ID, latitude, and longitude of the dataset. To achieve (2), it is necessary to extract the name of the place from the text. Geographical information is added to the database using various methods such as extracting data published in papers or as datasets, manually analyzing geospatial information, and extracting through natural language processing.