Current statistical machine translation (SMT) systems are trained on sentencealigned and word-aligned parallel text collected from various sources. Translation model parameters ar...
Spyros Matsoukas, Antti-Veikko I. Rosti, Bing Zhan...
This paper presents an attempt at building a large scale distributed composite language model that simultaneously accounts for local word lexical information, mid-range sentence s...
The translation quality and parsing efficiency are often disappointed when Rule based Machine Translation systems deal with long sentences. Due to the complicated syntactic structu...
In recent years, corpus based approaches to machine translation have become predominant, with Statistical Machine Translation (SMT) being the most actively progressing area. Succe...
This paper presents a method for improving phrase-based Statistical Machine Translation systems by enriching the original translation model with information derived from a multilin...