We address the problem of smoothing translation probabilities in a bilingual N-grambased statistical machine translation system. It is proposed to project the bilingual tuples ont...
Minimum Error Rate Training is the algorithm for log-linear model parameter training most used in state-of-the-art Statistical Machine Translation systems. In its original formula...
Virtual memoy page translation tables provide mappings from virtual to physical addresses. When the hardware controlled Tratmlation L.ookaside Buffers (TLBs) do not contain a tran...
The need for automated text evaluation is common to several AI disciplines. In this work, we explore the use of Machine Translation (MT) evaluation metrics for Textual Case Based R...
Ibrahim Adeyanju, Nirmalie Wiratunga, Robert Lothi...
Extant Statistical Machine Translation (SMT) systems are very complex softwares, which embed multiple layers of heuristics and embark very large numbers of numerical parameters. A...