Arabic Diacritization through Full Morphological Tagging

We present a diacritization system for written Arabic which is based on a lexical resource. It combines a tagger and a lexeme language model. It improves on the best results reported in the literature.

• Type: Conference
• Year: 2007
• Where: NAACL
• Authors: Nizar Habash, Owen Rambow
• Keywords: Computational Linguistics, Lexeme Language Model, NAACL 2007, Written Arabic