Publications
CONFERENCE (INTERNATIONAL) DisasterNeedFinder: Understanding the Information Needs in the Noto Earthquake
Kota Tsubouchi, Shuji Yamaguchi, Keijirou Saitou (Japan Broadcasting Corporation), Akihisa Soemori (NHK Global Media Servises), Masato Morita (NHK Global Media Servises), Shigeki Asou (Japan Broadcasting Corporation)
The 32nd ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM SIGSPATIAL 2024)
October 29, 2024
The Noto Peninsula Earthquake on January 1, 2024, in Ishikawa Prefecture, Japan, had a magnitude of 7.6, causing tsunamis, landslides, fires, and liquefaction. Over 200 people died, and the Oku-Noto region suffered significant damage. The earthquake disrupted transportation and complicated rescue operations, while utilities like electricity, water, and gas were severely affected, with many areas still not restored by June 2024. This study proposes the "DisasterNeedFinder" framework to address dynamic information needs during large-scale disasters. It integrates search history and location data, collecting both search queries and user location data to determine the geographic origin of the queries. Machine learning generates stopwords to offset media influence and identify regional characteristics. The framework defines information needs intensity based on query anomalies, effectively capturing real-time needs despite unstable access conditions and low user density. The framework, validated during the Noto Peninsula Earthquake, proved effective in real-time assessment of information needs, applicable to various large-scale disasters.
Paper :
DisasterNeedFinder: Understanding the Information Needs in the Noto Earthquake
(external link)