ML models predict next speaker in multi-party dialogues, with content-based deep learning performing best on large corpora while simpler speaker-only models suffice for small topic-focused ones.
Multiwoz - a large-scale multi-domain wizard-of-oz dataset for task-oriented dialogue modelling
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning Multi-Party Turn-Taking Models from Dialogue Logs
ML models predict next speaker in multi-party dialogues, with content-based deep learning performing best on large corpora while simpler speaker-only models suffice for small topic-focused ones.