n[lm-probs]= trigram unigram trigram-unigram
n[rates]= seg_tr_syl seg_tr_phn lex_syl lex_phn enrate
vrate nvrate mrate mfrate enmmfrate
mmfrate
n[positions] =
word_num_in_utterance word_num_in_turn
n[morebigrams] = bigram reverse-bigram reverse-trigram
center-trigram
npart-of-speech = syntactic part of speech (currently
only done for the word "to")
nwordX= word
number X in acoustically segmented `sentence'
ncanonical?=
can if canonical (pronlex) pronunciation, alt otherwise
ntrigram= p(word | previous two words)
nunigram= p(word)
ntrigram-unigram = difference between two
probabilities
nseg_tr_syl= transcribed syllable rate between closest
two pauses
nseg_tr_phn= transcribed phone rate between closest two
pauses
nlex_syl= lexical syllabic rate (i.e. as determined from
wd transcription)
nlex_phn= lexical phone rate (i.e. as determined from wd
transcription)