Syntax-enhanced n-gram-based SMT.
Josep Crego, José Bernardo Mariño Acebal · 2007
This paper addresses the problem of word re-ordering in statistical machine translation. We follow a word order monotonization strategy making use of syntax information (dependency parse tree) of the source language to build a set of automatically extracted reordering rules. The input sentence is extended to a graph built with reordering hypotheses, hence, allowing for a constrained search on the syntactically moti-vated reorderings. Experiments are reported on the BTEC cor-pus (Chinese to English task) Results are pre-sented regarding translation accuracy and com-putational efficiency, showing significant im-provements in translation quality at a reason-able computational cost. 1