Unveiling political polarization on Twitter: Machine learning and sentiment analysis in presidential elections

Abstract

The year 2024 stands out as a pivotal year marked by significant political transformations across the globe. Some countries, such as Mexico and the United States, could be deeply affected by political polarization and echo chambers. This study employed sentiment analysis and machine learning techniques to investigate political polarization on Twitter during the 2018 Mexican presidential election. The findings reveal that the winning candidate exhibited the highest level of polarization. This underscores the pivotal role of social media in elections. For some time now, social media platforms like Twitter have contributed to intensified political polarization and the creation of echo chambers. Further research is essential to understand the influence of polarization on voter decision-making and democratic procedures. Establishing ethical guidelines for using machine learning in policy analysis is critical to preserving the integrity of democratic processes while reaping the potential benefits of new technologies.

Keywords

• political polarization
• presidential elections
• artificial intelligence
• machine learning
• sentiment analysis