CAMERCulture-Aware Multilingual Emotion Recognition for Low-Resource Languages

This project investigates how cultural context shapes the expression and interpretation of emotions across languages, with a focus on low-resource and underrepresented communities. The goal is to develop culture-aware multilingual emotion recognition models that remain reliable when training data is scarce, labels are noisy, and emotion categories do not map cleanly across cultures. The expected outcome is a set of models, datasets, and evaluation tools that enable fairer and more accurate emotion-aware NLP for applications in education, public services, and human-centered conversational systems without marginalizing languages and cultures that are currently underserved.