ACSPRI Conferences, RC33 Eighth International Conference on Social Science Methodology

Font Size:  Small  Medium  Large

Exploring Computational Methodologies for Peoples' Behaviors at Disaster: Analyses on Interconnections of People at the Tohoku Quake in Japan

Kazuhiko Shibuya

Building: Law Building
Room: Breakout 5 - Law Building, Room 020
Date: 2012-07-11 03:30 PM – 05:00 PM
Last modified: 2012-07-03


The main goal of this paper was to deepen understandings for information seeking behavior by citizens, published documents that written by each stakeholder of the Fukushima Crisis and socially-oriented collective networking by voluntary participants at the Tohoku Disaster. Critically understandings on real-time analysis for attitudes of citizens, verifications on managements by stakeholders and the nature of collective networking by voluntary participants at disaster will be more necessary than ordinary daily life. So then, I conducted current studies by some computational methodologies such as web data analysis, text mining, and agent-based simulation.
First of all, this study unveiled behavior patterns of ordinary citizens for their safety after the Tohoku Earthquake (11th March, 2011) in Japan. One of the most important facts was that voluntary citizens used social media and they activated interconnecting among each other for supportive collaborations. Their behaviors were characterized as information seeking behavior using social media and collaborative tools. I critically verified their behaviors and communication patterns. So then, using various web data analysis tool and other methodologies, I found some evidences and implications of information seeking behavior by ordinary citizens in uncertain social conditions in the Tohoku Quake.
Secondly I intended to understand the nature of the Fukushima Crisis from massive published documents by means of computational ways. As my motivations on sociology of science, I conducted to analyze various types of documents that written by each stakeholder in Japan. In this case, I analyzed these data by text mining.
Thirdly I clarified a type of collective networking by peoples using an agent-based simulation. It means to understand the nature of collective networking and its properties of grass-root voluntarism in communities. This networking was considered as a kind of self-organizing patterns, and these patterns were mostly matched for actual observations in the Tohoku Quake.
And finally, taken together, I discussed the lessons that I learnt from the Tohoku Disaster.

Full Text: Full paper PDF