Deciphering Public Opinion of Nuclear Energy on Twitter
Authors:
- 1 July 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-8
Abstract
This paper explores nuclear energy-related Twitter discussions as a response to the 2011 Fukushima Nuclear Disaster and the 2017 Nobel Peace Prize won by the International Campaign to Abolish Nuclear Weapons. We have considered a total of 2 million tweets for these two events. In particular, we employed CNN, LSTM, and Bi-LSTM to investigate whether social media users are supportive or cynical about nuclear energy. Our AI algorithms have performed better for polarity detection (accuracy in the range of 90%) with respect to subjectivity detection (accuracy in the range of 75%). We also note that dominant aspects of supporting tweets revolve around concepts like clean energy, lower CO2 emission, and sustainable future. On the contrary, cynical users see nuclear energy as a threat to the environment, human life, and safety.Keywords
This publication has 34 references indexed in Scilit:
- Deep learningNature, 2015
- New Trends of Learning in Computational Intelligence [Guest Editorial]IEEE Computational Intelligence Magazine, 2015
- Modelling Public Sentiment in Twitter: Using Linguistic Patterns to Enhance Supervised Learning2015
- Social Media and the New Organization of Government CommunicationsThe American Review of Public Administration, 2014
- Word Polarity Disambiguation Using Bayesian Model and Opinion-Level FeaturesCognitive Computation, 2014
- Citizen–government collaboration on social media: The case of Twitter in the 2011 riots in EnglandGovernment Information Quarterly, 2014
- A Convolutional Neural Network for Modelling Sentences2014
- Coooolll: A Deep Learning System for Twitter Sentiment Classification2014
- Convolutional Neural Networks for Sentence Classification2014
- Long Short-Term MemoryNeural Computation, 1997