tech,6-1-N03-1001,bq This paper describes a method for <term> utterance classification </term> that does not require <term> manual
other,12-1-N03-1001,bq classification </term> that does not require <term> manual transcription </term> of <term> training data </term> . The
lr,15-1-N03-1001,bq <term> manual transcription </term> of <term> training data </term> . The method combines <term> domain
model,3-2-N03-1001,bq training data </term> . The method combines <term> domain independent acoustic models </term> with off-the-shelf <term> classifiers
tech,9-2-N03-1001,bq acoustic models </term> with off-the-shelf <term> classifiers </term> to give <term> utterance classification
measure(ment),12-2-N03-1001,bq off-the-shelf <term> classifiers </term> to give <term> utterance classification performance </term> that is surprisingly close to what
tech,26-2-N03-1001,bq can be achieved using conventional <term> word-trigram recognition </term> requiring <term> manual transcription
other,29-2-N03-1001,bq word-trigram recognition </term> requiring <term> manual transcription </term> . In our method , <term> unsupervised
tech,4-3-N03-1001,bq transcription </term> . In our method , <term> unsupervised training </term> is first used to train a <term> phone
model,12-3-N03-1001,bq training </term> is first used to train a <term> phone n-gram model </term> for a particular <term> domain </term>
other,18-3-N03-1001,bq n-gram model </term> for a particular <term> domain </term> ; the <term> output </term> of <term> recognition
other,21-3-N03-1001,bq particular <term> domain </term> ; the <term> output </term> of <term> recognition </term> with this
tech,23-3-N03-1001,bq domain </term> ; the <term> output </term> of <term> recognition </term> with this <term> model </term> is then
model,26-3-N03-1001,bq of <term> recognition </term> with this <term> model </term> is then passed to a <term> phone-string
tech,32-3-N03-1001,bq <term> model </term> is then passed to a <term> phone-string classifier </term> . The <term> classification accuracy
measure(ment),1-4-N03-1001,bq phone-string classifier </term> . The <term> classification accuracy </term> of the <term> method </term> is evaluated
tech,5-4-N03-1001,bq classification accuracy </term> of the <term> method </term> is evaluated on three different <term>
other,11-4-N03-1001,bq </term> is evaluated on three different <term> spoken language system domains </term> . Motivated by the success of <term>
skrij podrobnosti