measure(ment),3-1-I05-5003,bq from the <term> Web </term> . The task of <term> machine translation ( MT ) evaluation </term> is closely related to the task of
tech,16-1-I05-5003,bq </term> is closely related to the task of <term> sentence-level semantic equivalence classification </term> . This paper investigates the utility
measure(ment),8-2-I05-5003,bq investigates the utility of applying standard <term> MT evaluation methods ( BLEU , NIST , WER and PER ) </term> to building <term> classifiers </term>
tech,22-2-I05-5003,bq , WER and PER ) </term> to building <term> classifiers </term> to predict <term> semantic equivalence
other,25-2-I05-5003,bq <term> classifiers </term> to predict <term> semantic equivalence </term> and <term> entailment </term> . We also
other,28-2-I05-5003,bq <term> semantic equivalence </term> and <term> entailment </term> . We also introduce a novel <term>
tech,5-3-I05-5003,bq </term> . We also introduce a novel <term> classification method </term> based on <term> PER </term> which leverages
measure(ment),9-3-I05-5003,bq classification method </term> based on <term> PER </term> which leverages <term> part of speech
other,12-3-I05-5003,bq on <term> PER </term> which leverages <term> part of speech information </term> of the <term> words </term> contributing
other,18-3-I05-5003,bq of speech information </term> of the <term> words </term> contributing to the <term> word matches
other,22-3-I05-5003,bq <term> words </term> contributing to the <term> word matches and non-matches </term> in the <term> sentence </term> . Our
other,28-3-I05-5003,bq matches and non-matches </term> in the <term> sentence </term> . Our results show that <term> MT evaluation
measure(ment),4-4-I05-5003,bq sentence </term> . Our results show that <term> MT evaluation techniques </term> are able to produce useful <term> features
other,12-4-I05-5003,bq techniques </term> are able to produce useful <term> features </term> for <term> paraphrase classification
tech,14-4-I05-5003,bq produce useful <term> features </term> for <term> paraphrase classification </term> and to a lesser extent <term> entailment
tech,21-4-I05-5003,bq classification </term> and to a lesser extent <term> entailment </term> . Our <term> technique </term> gives
tech,1-5-I05-5003,bq extent <term> entailment </term> . Our <term> technique </term> gives a substantial improvement in
measure(ment),7-5-I05-5003,bq gives a substantial improvement in <term> paraphrase classification accuracy </term> over all of the other <term> models
tech,15-5-I05-5003,bq accuracy </term> over all of the other <term> models </term> used in the experiments . We propose
skrij podrobnosti