Public opinion sensing and trend analysis on social media: A study on nuclear power on twitter

Abstract

Due to their popularity and rapid propagation capability, social media have become important communication channel on public opinions because they work as a main virtual space to raise new public issues, discussion on these issues, and make opinion directions on them. With the recognition of the importance of social media, many enterprises and governments try to use social media as marketing tools or public opinion sensing tools. Nuclear power has become one of trending public issues in Korea after Fukushima nuclear disaster. Nuclear power is a double-edged sword because it is the most efficient way to generate electricity currently and it also has dangerous potential risk such as radiation leakage. Governments and policy makers need to monitor public opinion on nuclear power continuously because it can be changed depending on related events. In this study, we aim to propose an approach to sense public opinions on nuclear power and to analyze their trend based on opinion mining techniques. Also we propose a measure to trace the changes on directions of public opinion on nuclear power. To show the usefulness of the proposed approach, we had gathered tweets on nuclear power in Korean from 2009 to 2013. After classification of tweets from 2009 to 2011, sentimental dictionary including positive and negative terms had been constructed. The performance testing using tweets from 2012 to 2013 showed that the proposed approach can be applied in practice.