Practice and philosophy in machine learning: Evaluation and critical reflection of the model-dependency of machine-learning algorithms for hate speech classification
Year: 2025
Type: dissertation
Abstract: This study seeks to fine-tune two hate-speech detection models using the same dataset in order to uncover intrinsic differences that may affect how just they are. It includes both a technical analysis, which quantifies biases through an explainability method, and an ethical analysis. The ethical analysis explores the values embedded in the models' design, anticipates potential negative consequences within the theoretical context of deployment, and proposes ideal practices to reduce such issues.
Author: Sabater Vilchez, Blanca
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7112471617
