Understanding Social Media's Take on Climate Change through Large-Scale Analysis of Targeted Opinions and Emotions

摘要

Social media provides a powerful data source for researchers interested in understanding population-level behavior, having been successfully leveraged in a number of different application areas including disease prediction models, detecting civil unrest, and measuring public sentiment towards a given topic of interest within the public discourse. In this work, we present a study of a large collection of Twitter data centered on the social conversation around global climate change during the 2015 UN Climate Change Conference, held in Paris, France in December 2015. We first developed a mechanism for distinguishing between personal and non-personal accounts. We then analyzed demographics, emotion, and opinion dynamics over time and location in order to understand how the different user types converse around meaningful topics on social media. This methodology offers in-depth insight into the behavior and opinions around a topic where multiple distinct narratives are present involved, and lays the groundwork for future work in studying narratives in social media.