Analysis of Public Perception using Twitter in Nuclear Energy Technology

Abstract

This paper uses one of the most common social media networks, Twitter, to analyze trends in public perception about nuclear energy technology. Our model uses python scripts since it is a rich set of packages for Natural Language Processing (NLP).

Our methodology consists of three sections. In the first section we randomly scrape 1,000 tweets for “Nuclear Energy” using the Twitter Intelligence Tool (TWINT). In the second section we run Sentiment Analysis of each tweet using the Valence Aware Dictionary and sEntiment Reasoner (VADER). We created three labels: positive, negative, and neutral. The third section is part of the Exploratory Data Analysis (EDA) and counts the frequency of each of the three sentiment classes in form of word-clouds.

We believe that an analysis of Twitter trends can provide rich insights on public perception about nuclear energy. Understanding them well can be useful for creating good strategies for mass communications.