Visualization of SNS Communications Concerning Radiation Risk from the Fukushima Daiichi Power Plant

In this presentation, I visualize people's communications on a Social Networking Service (SNS) concerning the risks of radioactive pollution stemming from the Fukushima Daiichi power plant accident. The data consists of the communication log from a regional SNS that was launched by the residents of Iwaki City, Fukushima Prefecture, after the accident. It contains text from diaries and comments written by the SNS members and information about who made comments to whom. The co-occurrence network of words derived from the text data reveals the members' concerns about radiation risks. Comment relations are expressed as a form of social network that shows the structure of the human relationships on the SNS. The co-occurrence network of words and the comment network can be integrated and analyzed concurrently to show the development and interaction of the members' shared concerns and the social networks in the SNS.

The SNS members expressed various concerns, including radiation measurement, influence on health, food safety, and so on. Based on their concerns, they created clusters in the comment network. Though the founder of the SNS had the highest degree centrality, the mediator of the clusters was a different member who sometimes mediated members with opposing opinions. A member who had a scientific background occupied a leadership position in a particular cluster but did not play a role in mediation. This suggests that a sympathetic attitude is more important than scientific knowledge for risk communication among ordinary citizens.