A hybrid Noun Phrase translation system

We present a hybrid based Noun Phrase (NP) translator that combines rule-based transfer technique with a statistical n-gram language model for selecting the best translation. Noun Phrase is the dominating construct in natural language text and targeting it for focused processing increases effectiveness of language processing systems. Manipulation of Noun Phrases is an effective subtask in Statistical Machine Translation, Multilingual Information Retrieval and Information Extraction. In this work, we make use of knowledge about Arabic language morphology regarding the translation of Verbal Nouns (Masader) and Annexation Constructs.