The authors introduce the task of asking clarifying questions for open-domain information-seeking conversations, collect the Qulac dataset from TREC topics, and propose a retrieval framework that outperforms baselines with an oracle showing 170% P@1 gain.
Trippas, Damiano Spina, Lawrence Cavedon, Hideo Joho, and Mark Sanderson
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
-
Asking Clarifying Questions in Open-Domain Information-Seeking Conversations
The authors introduce the task of asking clarifying questions for open-domain information-seeking conversations, collect the Qulac dataset from TREC topics, and propose a retrieval framework that outperforms baselines with an oracle showing 170% P@1 gain.