Twitter Analysis of Public Acceptance between Seoul and Gori of Nuclear Power

Public acceptance is critical for sustaining nuclear power, and researchers have devised various methods to measure it (Stritar, 1996). Existing literatures suggest that rationality, emotion, knowledge of nuclear technology, trust, policy executor, and risk perception variables affect public acceptance. These attempts, however, have been limited to epistemological measurements using methods such as the Likert scale (Sjoberg and Drottz-Sjoberg, 2009; Slovic, 2000; Tanaka, 2004). Because such methods are standardized, it is difficult to reflect on emotions latent in individuals within the public. Moreover, surveys can be conducted only on people in a specific region and time interval, and it may be misleading to generalize the results to represent the attitude of the public. Because big data methods are based on unstructured data, which contain the live experiences/opinions, and are virtually real-time with almost no delay between the events of concern and the data collection, big data analysis allows real-time identification of relationships among different variables and their significance (Graham and Shelton, 2013). In this research, we propose big data analysis as a solution and attempt to identify the attitudes of the public on nuclear energy using big data analysis. To conclude, big data is a useful tool to measure the public acceptance of nuclear technology efficiently (i.e., saves cost, time, and effort of measurement and analysis), and this research was able to provide a case for using big data to analyze the public acceptance of nuclear technology. The trends and opinions of opinion leaders on SNSs should be monitored and responded to in real time. As demonstrated from the rejection of the extension of the Gori nuclear power plant, image and feeling are more important than the performance of the safety technology on the operation of a nuclear power plant. Because Korea has many aging reactors, evaluation of projects to extend the operation of old reactors and build new reactors will take place in the near future. In this context, it is crucial to collect and analyze data regarding the image of nuclear power. The big data analysis will allow the nuclear industry and the government to proactively respond to the public, which will lead to rational decisions and interaction with the public. Evaluation and prediction of nuclear policies will be made more reliable by using the big data methodologies and the results of this research. Nevertheless, there are certain limitations. First, this research did not analyze the individuals directly, but their group in an SNS. Applying the results to the individuals directly may create an ecological fallacy (Schwartz, 1994). Second, using only Twitter data could limit the scope of this research (Li et al., 2012). Nevertheless, the research collected ample datasets for the analysis, and the overall trend of opinions is not expected to differ even in other SNSs