Dependency Parsing of Modern Standard Arabic with Lexical and Inflectional Features

Abstract

We explore the contribution of lexical and inflectional morphology features to dependency parsing of Arabic, a morphologically rich language. Using controlled experiments, we find that definiteness, person, number, gender, and undiacritized lemma are most helpful for parsing on automatically tagged (predicted) input. We contrast it with using gold input. We further contrast the contribution of form-based and functional features, and show that functional features for gender and number (e.g., “broken plurals”) and the related rationality (“humanness”) feature improve over form-based features. We show that training on a combination of predicted and gold features improves over the alternatives. We examine the contribution of these features – some of which only recently introduced for Arabic NLP – in two transition-based parsers: MaltParser and Easy-First Parser.