Quantitative visualization about differences between scientists concerned nature disasters and historic events

Abstract

How to process massive historic natural disasters events is a great challenge to recognize patterns. And more and more scientific research data provides a new source of nature disasters. In this study, the biclustering method is used to categorize the scientists concerned natural disasters and historic events. Cartograms, one kind of transformed maps, are created to highlight numbers of publications and events in a country. Reaction index (RI) is introduced to evaluate the difference between scientists concerned nature disasters and historic events. The results show that biclustering is a useful method to categorize data with high volumes and dimensions. Cartograms could represent conceptual patterns that are difficult to be displayed in regular maps. Analysis indicates that earthquakes and landslides attract relatively more concerns from scientists in the north hemisphere; floods are more focused by scientists in the south hemisphere. Although droughts are not significant in the cartogram of historic events, they obtain attentions from scientists of inland as well. The distribution of RIs shows that more scientists need to put more efforts in dealing with natural disasters, especially in Indonesia and Philippines.