PublicPulse: A Social Media Analysis Framework for Citizen Reactions to Government Actions

Social media has grown significantly, attracting more people who spend more time on these platforms and share vast amounts of information about their daily lives. Consequently, governments use such social media to disseminate information about their achievements, updated laws, and decisions. Simultaneously, citizens find social media platforms, such as Facebook, X (previously known as Twitter), and Instagram, appealing virtual spaces for expressing their thoughts and opinions. In brief, social media is becoming an increasingly vital channel of communication between governments and citizens. As a result, analyzing social media data to understand and predict public sentiment in the context of government-related events has become increasingly important. While sentiment analysis methods can be employed to mine sentiment from individual posts, estimating collective public sentiment about an event remains challenging. Specifically, sentiment analysis methods are associated with inaccuracies, and a group of individuals may express contradicting opinions about the same event. To address these challenges, we propose a novel framework named “PublicPulse”, designed to analyze social media posts and their associated metadata. PublicPulse incorporates a set of sentiment analysis methods along with statistical analysis to generate a public sentiment score for a set of posts related to a specific event. The framework was empirically evaluated using several events of interest to the government in Jordan from 2018 to 2022 and demonstrated accurate results when compared with the historical context of governmental actions.