tech,8-2-P99-1022,ak new <term> cluster generation </term> , <term> hierarchical smoothing </term> and <term> adaptive topic-probability
tech,11-2-P99-1022,ak <term> hierarchical smoothing </term> and <term> adaptive topic-probability estimation techniques </term> . These combined models help capture
tech,10-1-P99-1022,ak method of generating and applying <term> hierarchical , dynamic topic-based language models </term> . It proposes and evaluates new <term>
other,5-3-P99-1022,ak These combined models help capture <term> long-distance lexical dependencies </term> . Experiments on the <term> Broadcast
measure(ment),10-4-P99-1022,ak </term> show significant improvement in <term> perplexity </term> ( 10.5 % overall and 33.5 % on <term>
tech,5-2-P99-1022,ak </term> . It proposes and evaluates new <term> cluster generation </term> , <term> hierarchical smoothing </term>
other,19-4-P99-1022,ak </term> ( 10.5 % overall and 33.5 % on <term> target vocabulary </term> ) . In this paper we describe a systematic
lr-prod,3-4-P99-1022,ak dependencies </term> . Experiments on the <term> Broadcast News corpus </term> show significant improvement in <term>
hide detail