| lr,15-1-N03-1001,bq | <term> manual transcription </term> of <term> | training data | </term> . The method combines <term> domain | #2220 This paper describes a method for utterance classification that does not require manual transcription of training data . |
