Advancing arabic dialect detection with hybrid stacked transformer models

The rapid expansion of dialectally unique Arabic material on social media and the internet highlights how important it is to categorize dialects accurately to maximize a variety of Natural Language Processing (NLP) applications. The improvement in classification performance highlights the wider variety of linguistic variables that the model can capture, providing a reliable solution for precise Arabic dialect recognition and improving the efficacy of NLP applications. Recent advances in deep learning (DL) models have shown promise in overcoming potential challenges in identifying Arabic dialects. In this paper, we propose a novel stacking model based on two transformer models, i.e., Bert-Base-Arabertv02 and Dialectal-Arabic-XLM-R-Base, to enhance the classification of dialectal Arabic. The proposed model consists of two levels, including base models and meta-learners. In the proposed model, Level 1 generates class probabilities from two transformer models for training and testing sets, which are then used in Level 2 to train and evaluate a meta-learner. The stacking model compares various models, including long-short-term memory (LSTM), gated recurrent units (GRU), convolutional neural network (CNN), and two transformer models using different word embedding. The results show that the stacking model combination of two models archives outperformance over single-model approaches due to capturing a broader range of linguistic features, which leads to better generalization across different forms of Arabic. The proposed model is evaluated based on the performance of IADD and Shami. For Shami, the Stacking-Transformer achieves the highest performance in all rates compared to other models with 89.73 accuracy, 89.596 precision, 89.73 recall, and 89.574 F1-score. For IADD, the Stacking-Transformer achieves the highest performance in all rates compared to other models with 93.062 accuracy, 93.368 precision, 93.062 recall, and 93.184 F1 score. The improvement in classification performance highlights the wider variety of linguistic variables that the model can capture, providing a reliable solution for precise Arabic dialect recognition and improving the efficacy of NLP applications.