What AI says about lorenzofamiglini/Irony-Sarcasm-Detection-Task

The Irony-Sarcasm-Detection-Task project aims to detect irony and sarcasm in natural language processing. It analyzes traditional machine learning features and hyperparameter optimization, applying ensemble methods such as Bayesian Model Averaging. It also experiments with deep learning approaches, including DeepMoji and RoBERTa-based transformer models, using transfer learning and attention-based architectures with Bert Tweet and DeepMoji as feature extractors, combined in ensembles to improve performance.