Transformer's Models To Identify Twitter Users' Voting Intent Prior To A Presidential Election

Abstract

This research evaluates the use of Transformer-based models, such as BERT and RoBERTa, to identify voting intention on social networks during the period leading up to presidential elections in Ecuador. Relevant posts from the platform “X” were collected and processed using natural language processing techniques. Results demonstrate the effectiveness of these models in detecting political behaviour patterns, providing a useful tool for electoral analysis and strategic campaign planning.