Hanada Taha-Thomure

A Large and Balanced Corpus for Fine-grained Arabic Readability Assessment

This paper introduces the Balanced Arabic Readability Evaluation Corpus (BAREC), a large-scale, fine-grained dataset for Arabic readability assessment. BAREC consists of 69,441 sentences spanning 1+ million words, carefully curated to cover 19 readability levels, from kindergarten to postgraduate co…