Handling unknown words in statistical latent-variable parsing models for Arabic, English and French

Abstract

This paper presents a study of the impact of using simple and complex morphological clues to improve the classification of rare and unknown words for parsing. We compare this approach to a language-independent technique often used in parsers which is based solely on word frequencies. This study is applied to three languages that exhibit different levels of morphological expressiveness: Arabic, French and English. We integrate information about Arabic affixes and morphotactics into a PCFG-LA parser and obtain stateof-the-art accuracy. We also show that these morphological clues can be learnt automatically from an annotated corpus.