Answers are grounded in this paper and cite their sources.
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.
No takes yet. Share an insight, caveat, or question.
Habash et al. (2007) studied this question.