|
P | N
|
lexical-cohesion and conversational features |
2 |
|
|
P | N
|
ASR output |
2 |
|
|
P | N
|
transcription errors |
1 |
|
|
P | N
|
subtopic boundaries |
1 |
|
|
P | N
|
spoken multiparty dialogue |
1 |
|
|
P | N
|
segment boundaries |
1 |
|
|
P | N
|
predicting top-level topic shifts |
1 |
|
|
P | N
|
predicting top-level boundaries |
1 |
|
|
P | N
|
predicting top-level and predicting subtopic boundaries |
1 |
|
|
P | N
|
performance |
1 |
|
|
P | N
|
overlapping speech |
1 |
|
|
P | N
|
machine learning approach |
1 |
|
|
P | N
|
lexical cohesion-based approach |
1 |
|
|
P | N
|
identifying subtopic boundaries |
1 |
|
|
P | N
|
human transcription |
1 |
|
|
P | N
|
features |
1 |
|
|
P | N
|
cue phrases |
1 |
|
|
P | N
|
conversational cues |
1 |
|
