The Named Entity Recognition (NER) task aims to identify and categorize proper and important nouns in a text. This Natural Language Processing task proved to be challenging for languages with a rich morphology such as the Arabic language. In this paper, We introduce a new named entity recognizer for Arabic. This recognizer is based on Conditional Random Fields (CRF) and an optimized feature set that combines contextual, lexical, morphological and gazetteers features. Our system outperforms the state-of-the-art Arabic NER systems with a F-measure of 93.5% when applied to ANERcorp standard dataset.
展开▼
