A morphologically annotated dataset and comprehensive evaluation of morphological analyzers for Orani Arabic (MADOran)

The Orani dialect of Arabic is an under-resourced Algerian language variety which also suffers from a lack of systematic evaluation of morphological analyzers. This study attempts to fill a critical resource gap for dialectal Arabic NLP, linguistic research, and educational applications. It presents MADOran, a morphologically annotated corpus for the Orani dialect of Arabic (ORN), together with a systematic evaluation of morphological analyzers across conventional, deep, and transformer-based approaches. The dataset contains 30,919 words drawn from: written texts (41%) on topics such as college life, culture, history, humor, politics, and traditions; and spoken material (59%) from storytelling, music, and everyday exchanges. Each token received manual annotation for part-of-speech, root, pattern, and glosses in English and French, using the DIWAN tool and the annotation guidelines adapted from the Dynamic Arabella Corpus to suit dialectal properties. MADOran supports such dialect-focused NLP tasks as dialect identification, translation, generation, and recognition, while enabling comparison of ORN and Modern Standard Arabic (MSA) features in linguistic inquiry. It also aids lexicographic work and Arabic pedagogy. The resource observes FAIR principles, which ensure broad reusability. Experiments demonstrate its value for three-head morphological analysis (POS tagging, root extraction, and pattern recognition). A CRF baseline reached 88.54% accuracy on POS, 87.43% on roots, and 86.17% on patterns. Fine-tuning QARiB yielded the highest POS accuracy (90.04%), with a statistically significant relative gain of + 0.88% over the baseline.