other,7-1-P05-1032,bq In this paper we describe a novel <term> data structure </term> for <term> phrase-based statistical
tech,10-1-P05-1032,bq novel <term> data structure </term> for <term> phrase-based statistical machine translation </term> which allows for the <term> retrieval
tech,18-1-P05-1032,bq translation </term> which allows for the <term> retrieval </term> of arbitrarily long <term> phrases </term>
other,22-1-P05-1032,bq retrieval </term> of arbitrarily long <term> phrases </term> while simultaneously using less <term>
other,27-1-P05-1032,bq </term> while simultaneously using less <term> memory </term> than is required by current <term>
tech,33-1-P05-1032,bq </term> than is required by current <term> decoder </term> implementations . We detail the <term>
other,3-2-P05-1032,bq </term> implementations . We detail the <term> computational complexity </term> and <term> average retrieval times </term>
measure(ment),6-2-P05-1032,bq computational complexity </term> and <term> average retrieval times </term> for looking up <term> phrase translations
other,12-2-P05-1032,bq retrieval times </term> for looking up <term> phrase translations </term> in our <term> suffix array-based data
other,16-2-P05-1032,bq <term> phrase translations </term> in our <term> suffix array-based data structure </term> . We show how <term> sampling </term>
tech,3-3-P05-1032,bq data structure </term> . We show how <term> sampling </term> can be used to reduce the <term> retrieval
measure(ment),10-3-P05-1032,bq sampling </term> can be used to reduce the <term> retrieval time </term> by orders of magnitude with no loss
measure(ment),20-3-P05-1032,bq orders of magnitude with no loss in <term> translation quality </term> . We describe a novel <term> approach
skrij podrobnosti