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