Mining climate conversations on social media has largely focused on identifying stance (pro, anti, or neutral), detecting misinformation, and analyzing sentiment and emotions. However, all pro-climate change posts are not monolithic; some emphasize scientific discoveries and facts (science-centric), while others propose solutions and actions (action-centric). Distinguishing between action and science posts can inform effective messaging and solutions to combat climate change. This paper leverages three climate-focused subreddits to explore the differences between science and action posts with respect to user engagement, sentiment analysis and classification potential. The findings reveal that engagement metrics differ significantly between action and science posts and both science and action posts show nearly balanced sentiment. Extensive experimentation with various machine learning models demonstrates that science and action posts can be distinguished with high performance. We explain these findings by considering the attitudes of the citizens, and hence, contribute to the body of knowledge through a nuanced understanding of online discourse on climate change. Furthermore, our work highlights the strengths and limitations of various NLP techniques in distinguishing between science and action-oriented climate messaging.

In the early period of the COVID-19 pandemic, Michigan became a focal point of socio-political unrest, which also led to a parallel surge in Twitter activity. This study applies social network analysis (SNA) to 15,681 unique tweets related to Michigan protests to unravel the complex user interactions and conversational themes. SNA explores community structures, key figures, and prevailing narratives; connecting the physical protests to their online counterparts. An analysis of user interactions reveals many distinct, sparsely connected communities, accentuated by pivotal roles of political figures like Governor Whitmer and former President Trump. The influence of alternative media sources in driving and polarizing these conversations is also evident. An analysis of hashtag co-occurrences also reveals many distinct groups that probably share smaller subsets of overlapping interests within the broad realm of issues germane to the protests. Taken together, this suggests that users coalesce into polarized echo chambers, each dedicated to a few coherent issues; and concerns of one may be inherently at odds with others. The encouraging aspect, however, is that both user and hash tag networks carry the capacity for communication and exchange of ideas across these echo chambers. This capacity can be exploited to break down the polarization and cultivate empathy among disparate groups through exchange of perspectives. These findings can thus be leveraged to develop effective communication and sharing strategies, especially during socially turbulent situations.

Hateful and offensive speech on online social media platforms has been exacerbated by the turbulent and chaotic circumstances brought on by the coronavirus pandemic. A particularly contentious issue was lockdown orders issued by state governments designed to keep citizens safe by controlling the spread of the virus. To compel the government to relax these orders and restore normalcy, antilockdown protests were organized in many states. The economic, ideological, political, and health concerns related to the lockdowns and the associated protests were debated vigorously on social media platforms, many times using offensive content. Detecting such insulting and humiliating content is especially important during tumultuous times, when tensions are high, because such expressions online can quickly precipitate violence in the physical world. This paper presents an approach to detect hateful and offensive content from Twitter feeds collected after anti-lockdown protests in Lansing, Michigan. These tweets were labeled using a comprehensive definition of what constitutes offensive content based on its potential to trigger and incite people. Linguistic and auxiliary features were extracted from these labeled tweets. These features were further processed through feature selection and dimensionality reduction techniques. The preprocessed feature set was used to train machine learning models, which detect offensive content with an accuracy of around 84%. Our approach demonstrates the feasibility of identifying and tagging offensive content in politically motivated situations, even when such speech is dominated by contextual and circumstantial information. It can thus be used to mitigate the damage caused by widespread dissemination of offensive content.