Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/151651
Title: Diseño y desarrollo de aplicación para la detección de señales de alerta temprana de riesgo de suicidio mediante mensajes de texto de redes sociales y técnicas PLN
Author: Macias Martinez, Ignacio
Tutor: Divorra Vallhonrat, Teresa
Others: Merino Arranz, David
Abstract: Early detection of suicide risk through text analysis is favorable for mental health intervention and prevention. This project explores the application of Data Science and Artificial Intelligence to develop a system capable of identifying signs of suicidal risk in digital messages. Throughout the work, several key stages have been followed: 1. Exhaustive study of Natural Language Processing (NLP) and machine learning algorithms, with a focus on adapting these methods for the analysis of suicide-related texts. 2. Collection and processing of textual data through open sources, using relevant datasets to train and validate the models. 3. Implementation and evaluation of multiple predictive models, including neural networks, to classify texts into risk categories. 4. Development of a practical application using Streamlit, which incorporates the most effective model to enable real-time analysis of texts. 5. Analysis of the accuracy, reliability and applicability of the system in real contexts. The results obtained demonstrate a high accuracy in the identification of texts with suicidal risk content, validating the effectiveness of the PLN and machine learning approach in the field of mental health.
Keywords: procesamiento del lenguaje natural (PLN)
suicidio
detección de señales
Document type: info:eu-repo/semantics/bachelorThesis
Issue Date: 16-Jan-2024
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Bachelor thesis, research projects, etc.

Files in This Item:
File Description SizeFormat 
imaciasmTFG0124memoria.pdfMemoria del TFG1,71 MBAdobe PDFThumbnail
View/Open
Share:
Export:
View statistics

This item is licensed under aCreative Commons License Creative Commons