!MSA at BAREC Shared Task 2025: Ensembling Arabic Transformers for Readability Assessment
This addresses readability assessment for Arabic language processing, representing an incremental improvement through ensemble techniques and data augmentation.
The researchers tackled the problem of fine-grained Arabic readability assessment by creating a confidence-weighted ensemble of four transformer models with diverse loss functions, achieving first place in all six tracks of the BAREC 2025 Shared Task with 87.5% QWK at sentence level and 87.4% at document level.
We present MSAs winning system for the BAREC 2025 Shared Task on fine-grained Arabic readability assessment, achieving first place in six of six tracks. Our approach is a confidence-weighted ensemble of four complementary transformer models (AraBERTv2, AraELECTRA, MARBERT, and CAMeLBERT) each fine-tuned with distinct loss functions to capture diverse readability signals. To tackle severe class imbalance and data scarcity, we applied weighted training, advanced preprocessing, SAMER corpus relabeling with our strongest model, and synthetic data generation via Gemini 2.5 Flash, adding about 10,000 rare-level samples. A targeted post-processing step corrected prediction distribution skew, delivering a 6.3 percent Quadratic Weighted Kappa (QWK) gain. Our system reached 87.5 percent QWK at the sentence level and 87.4 percent at the document level, demonstrating the power of model and loss diversity, confidence-informed fusion, and intelligent augmentation for robust Arabic readability prediction.