Enhancing Bilingual Sarcasm Detection with an Advanced Ensemble Model: A Dual-Language Deep Learning Framework
Syed Nakibul Islam, Fahad Siddique Faisal, Md. Farhad Hossain, Md. Azad Hossain · 2024
Sarcasm is a type of verbal irony in which the literal expression of a statement contradicts its intended meaning. Sarcasm detection is crucial for interpreting web portal comments, customer feedback, and social media interactions, as misinterpretation can distort sentiment analysis. It improves the ability of automated systems to process opinions, reviews, and dialogues in web series or ordinary communication, thereby ensuring a more nuanced comprehension of user intent. This study has proposed a technique for Bilingual Sarcasm Detection that has used an ensemble model, integrating Both BiLSTM and GRU to detect sarcasm from both English and Bengali text. This study provides valuable insights into the effectiveness of Bi-LSTM and GRU in detecting subtle and context-dependent sarcastic expressions in both Bengali and English Languages. The proposed ensemble model utilizing word embeddings demonstrated superior performance compared to the other deep learning models achieving approximately 97% accuracy on the news headline dataset and 89% on the BanglaSARC dataset.