About

My current research focuses on the development of methodologies for the automatic identification and analysis of emotions in long, complex texts, with a particular emphasis on historical documents and other narrative sources. Building on prior work that seeks to move beyond simple sentiment polarity, I am developing models capable of detecting a richer and predefined set of emotions. This involves integrating techniques from automatic text analysis, sentiment analysis, and natural language processing to capture subtle affective cues, contextual patterns, and shifts in emotional tone throughout extended texts. The proposed methodology is evaluated using expert-annotated corpora to ensure precision and interpretability.
In addition to this central research line, I am currently directing several theses that reflect broader trends in artificial intelligence and the application of machine learning to real-world predictive and classification tasks:

• Machine learning for federal elections prediction, focusing on the construction of predictive models that incorporate historical electoral |data, demographic variables, and temporal patterns to estimate electoral outcomes with higher accuracy.


• Prediction of the amount of claims for mid-range vehicles using machine learning techniques, where students develop regression and forecasting models to identify risk factors and estimate expected claims in the automotive insurance sector.


• Automatic classification of digital messages aimed at middle childhood using NLP models, which explores how natural language processing and text classification techniques can identify, categorize, or filter content designed for children in the middle-childhood age range.

Together with my work on emotion detection in long texts, these thesis projects reflect an overarching research agenda that leverages machine learning and NLP to address complex analytical challenges across domains such as historical text interpretation, public decision-making, risk modeling, and digital communication. This integrated approach aims to produce both methodological advances and practical tools capable of supporting informed analysis in diverse applied contexts.

Resume

Education

Publications

Conferences