Domain adaptation is a major challenge for neural machine translation (NMT). Given unknown words or new domains, NMT systems tend to generate fluent translations at the expense of adequacy. We present a stack-based lattice search algorithm for NMT and show that constraining its search space with lattices generated by phrase-based machine translation (PBMT) improves robustness. We report consist...

throughout history several translators with different motives and styles have tried to translate the holy koran. the study of their goals and styles/ their strong and weak points can be a way to present improved translations of the holy koran in the future. focusing on the hamd sura, this article studies three well-known german translators of koran (salomon schweigger, david friedrich megerlin,...

In the Indian Holy Book – THE UPANISADS (800 B.C.), man has been described as “the son of the Sublime” (“Amritasya Putra”). It is recorded in the first chapter of the Book of Genesis (First Book of Holy Hebrew Bible [The Old Testament (c. 200 B.C.)] that God made man in his own image after his likeness as the crowning glory of his creation. The Holy Koran (c. 630 A.D.) says : “God sends down no...

t in islam world, writing and scribing koran has been always a very important issue. professional  scribers and calligraphers of koran were in the acme point of this process. scribing diversity as  well as different styles of writing divine words does not contradict with uprightness in recording  koran words so that in addition to observing this principle, great calligraphers have tried to  scr...