Project title
Cross-lingual Emotion Detection Through Text Using mBERT and Bi-LSTM Models.
Project aim
This project aims to enhance cross-lingual emotion detection capabilities in English and Spanish texts using advanced Natural Language Processing (NLP) techniques.
Project outline
Utilising the mBERT and Bi-LSTM models, the study explored both multilingual and cross-lingual approaches to understand and classify emotional expressions across linguistic boundaries.
The project leveraged the SemEval 2018 dataset and focused on improving interpretability and accuracy through Local Interpretable Model-agnostic Explanations (LIME).
Key findings emphasised the significance of optimised models, comprehensive datasets, and cultural context understanding to enhance emotion detection accuracy.
This research contributes to advancing human-computer interaction by improving how machines recognise and interpret human emotions across different languages.
Images
