African American

Research on African American English (AAE) focuses on mitigating biases in natural language processing (NLP) and automatic speech recognition (ASR) systems that stem from underrepresentation of AAE in training data. Current efforts utilize various machine learning models, including self-supervised learning models and XGBoost, to improve ASR accuracy and address biases in sentiment and toxicity analysis, often incorporating acoustic features, linguistic analysis, and data augmentation techniques. These advancements are crucial for ensuring equitable access to technology and reducing the perpetuation of societal biases embedded within AI systems.