Las redes sociales como factor de predicción de resultados electorales en campañas presidenciales

Abstract

The content generated in the interaction with social networks, by its nature, needs to be analyzed and explored through the application of techniques for the classification of feelings based on lexical and machine learning approaches, in order to find patterns or trends in people's behavior, which help candidates and political parties to measure the degree of acceptance and the correct direction of their proposals. This research project was able to determine that the content generated by the user on the social network Twitter, can be taken into account as a valid measurement factor to predict election results in presidential campaigns. The case study corresponds to the presidential elections of Colombia for the year 2018, in which 6,057,839 of tweets generated in Spanish were analyzed.