The research question that leads this effort is: How do Twitter messages impact the polarization of emotions in the presidential election in Mexico in 2018? We collected 6,515 tweets posted by the five candidates that stood for the presidency of Mexico in 2018. We developed a methodology based on artificial intelligence techniques using machine learning and classifiers to predict the emotions of each candidate and the effects of those emotions on the other candidates. The results show that the candidate with the greatest polarization in the generated emotions won the electoral contest and that the emotions generated in Twitter affected other candidates.