Detección de depresión a partir de comentarios escritos en español usando Procesamiento del Lenguaje Natural
Fecha
2024-03-05
Autores
Título de la revista
ISSN de la revista
Título del volumen
Editor
Jaén: Universidad de Jaén
Resumen
Con este Trabajo Fin de Grado se pretende aprovechar los avances en el campo del Procesamiento del Lenguaje Natural
para abordar la detección de indicios de depresión. Para ello, se han creado dos sistemas de clasificación automática. En
concreto, se ha desarrollado un sistema para participar en la Tarea 2.a: "Detección de depresión" de MentalRiskES, una
novedosa tarea para la identificación temprana de riesgo de trastornos mentales en comentarios en español, organizada en
el marco de la campaña de evaluación IberLEF. Aunque la tarea es sobre comentarios en español, el sistema que
presentamos es en inglés y aplica una traducción de los textos español-inglés, por lo que se ha analizado dicho sistema y
se ha realizado una modificación de las variables del modelo al español para usar los textos directamente en español.
Además, se ha realizado un análisis de resultados para evaluar el desempeño de ambos modelos.
This Final Degree Project aims to take advantage of the advances in the Natural Language Processing field to address the detection of signs of depression. To this end, two automatic classification systems have been developed. Specifically, one of them was developed to participate in Task 2.a: “Depression Detection” in MentalRiskES, a novel task for the early identification of risk of mental disorders in Spanish comments, organized within the framework of the IberLEF evaluation campaign. Although the task is about comments in Spanish, the system we present is in English and applies a translation of the Spanish-English texts, so this system has been analysed and a modification of the variables of the model into Spanish has been carried out in order to use the texts directly in Spanish. In addition, an analysis of the results has been carried out to evaluate the performance of both models.
This Final Degree Project aims to take advantage of the advances in the Natural Language Processing field to address the detection of signs of depression. To this end, two automatic classification systems have been developed. Specifically, one of them was developed to participate in Task 2.a: “Depression Detection” in MentalRiskES, a novel task for the early identification of risk of mental disorders in Spanish comments, organized within the framework of the IberLEF evaluation campaign. Although the task is about comments in Spanish, the system we present is in English and applies a translation of the Spanish-English texts, so this system has been analysed and a modification of the variables of the model into Spanish has been carried out in order to use the texts directly in Spanish. In addition, an analysis of the results has been carried out to evaluate the performance of both models.
Descripción
Palabras clave
Tratamiento Inteligente de la Información