Bibliographic citations
Carhuancho, A., Nuñez, S. (2024). NoHateS: Modelo de IA para la detección automática de contenido de odio en plataformas de interacción social [Tesis, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/676189
Carhuancho, A., Nuñez, S. NoHateS: Modelo de IA para la detección automática de contenido de odio en plataformas de interacción social [Tesis]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2024. http://hdl.handle.net/10757/676189
@misc{renati/892806,
title = "NoHateS: Modelo de IA para la detección automática de contenido de odio en plataformas de interacción social",
author = "Nuñez Lazo, Sergio Antonio",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2024"
}
Hate speech detection is a challenging task, especially in the context of real-time monitoring on the internet. Manual detection is both exhausting and impractical due to the high volume and frequency of online data. This thesis proposes a system called NoHateS. This system is made of multiple components, the main one is BETO-CNN, a Transformers-based model trained on a Spanish corpus, which is designed to actually detect whether a text contains hate speech or not. The second component is developed to ensure accessibility. This includes an API to allow seamless integration of the model into various applications, and a Discord Bot developed for easy manipulation of the aforementioned API in order to help users detect hate speech in text channels. This thesis also includes tests with imbalanced data and applies data augmentation in order to deal with it and make more robust models. The results demonstrate the effectiveness of NoHateS in detecting hate speech and provide recommendations for future research in this domain.
This item is licensed under a Creative Commons License