Neural Machine Translation for Arabic Language

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
The British University in Dubai (BUiD)
Translating the Arabic Language into other languages engenders multiple linguistic problems, as no two languages can match, either in the meaning given to the conforming symbols or in the ways in which such symbols are arranged in phrases and sentences. Lexical, syntactic and semantic problems arise when translating the meaning of Arabic words into English. Machine translation (MT) into morphologically rich languages (MRL) poses many challenges, from handling a complex and rich vocabulary, to designing adequate MT metrics that take morphology into consideration. The task of recognizing and generating paraphrases is an essential component in many Arabic natural language processing (NLP) applications. A well-established machine translation approach for automatically extracting paraphrases, leverages bilingual corpora to find the equivalent meaning of phrases in a single language, is performed by "pivoting" over a shared translation in another language. Neural machine translation has recently become a viable alternative approach to the more widely-used statistical machine translation. In this thesis, we revisit bilingual pivoting in the context of neural machine translation and present a paraphrasing model based mainly on neural networks. The thesis we present also, highlights the key challenges for Arabic language translation into English, and Arabic. Experimental results across datasets confirm that neural paraphrases significantly outperform those obtained with statistical machine translation, and indicate high similarity correlation between our model and human translation, making our model attractive for real-world deployment.
Neural networks (Computer science)., machine translation, deep learning, Arabic language, Arabic natural language processing (NLP)