tech,4-1-H05-1095,bq performance </term> . This paper presents a <term> phrase-based statistical machine translation method </term> , based on <term> non-contiguous phrases
other,12-1-H05-1095,bq translation method </term> , based on <term> non-contiguous phrases </term> , i.e. <term> phrases </term> with gaps
other,16-1-H05-1095,bq non-contiguous phrases </term> , i.e. <term> phrases </term> with gaps . A <term> method </term> for
tech,1-2-H05-1095,bq i.e. <term> phrases </term> with gaps . A <term> method </term> for producing such <term> phrases </term>
other,5-2-H05-1095,bq <term> method </term> for producing such <term> phrases </term> from a <term> word-aligned corpora </term>
lr,8-2-H05-1095,bq producing such <term> phrases </term> from a <term> word-aligned corpora </term> is proposed . A <term> statistical
tech,1-3-H05-1095,bq word-aligned corpora </term> is proposed . A <term> statistical translation model </term> is also presented that deals such
other,10-3-H05-1095,bq </term> is also presented that deals such <term> phrases </term> , as well as a <term> training method
tech,16-3-H05-1095,bq <term> phrases </term> , as well as a <term> training method </term> based on the maximization of <term>
measure(ment),23-3-H05-1095,bq </term> based on the maximization of <term> translation accuracy </term> , as measured with the <term> NIST
measure(ment),30-3-H05-1095,bq accuracy </term> , as measured with the <term> NIST evaluation metric </term> . <term> Translations </term> are produced
other,0-4-H05-1095,bq <term> NIST evaluation metric </term> . <term> Translations </term> are produced by means of a <term> beam-search
tech,7-4-H05-1095,bq Translations </term> are produced by means of a <term> beam-search decoder </term> . Experimental results are presented
tech,10-5-H05-1095,bq that demonstrate how the proposed <term> method </term> allows to better generalize from
lr,17-5-H05-1095,bq allows to better generalize from the <term> training data </term> . This paper investigates some <term>
skrij podrobnosti