A Predictive Study of the 2024 Presidential Elections

A predictive study was conducted to examine user opinions expressed on the social network YouTube in relation to the 2024 presidential elections in Mexico. The methodology applied natural language processing techniques together with supervised classification algorithms for the purpose of electoral estimation. The procedure began with the systematic extraction of YouTube comments through the analysis of hashtags associated with the presidential candidates Claudia and Xóchitl. To this end, a download schedule with varied time slots was designed in order to obtain a stochastic and representative sample. A team of six researchers participated in the data collection process to help ensure heterogeneity and randomness in the dataset. The collected data were modelled using the Support Vector Machine algorithm, Naive Bayes, and linear regression to estimate trends. The results suggested that Claudia Sheinbaum would be the election winner, a prediction that was found to be consistent with the official election results. Smart citations: https://scite.ai/reports/10.61467/2007.1558.2026.v17i1.1196