These are the datasets for paper: Continuity of Topic, Interaction, and Query: Learning to Quote in Online Conversations.
There is a new work on these datasets, "Quotation Recommendation and Interpretation Based on Transformation from Queries to Quotations". The code is available at link.
The datasets are only for research use.
| # of quotes | 1053 | 1111 |
| Avg len of quotes | 4.0 | 10.1 |
| |Voc| of quotes | 1251 | 4111 |
| # of convs | 19081 | 44539 |
| Avg # of turns per conv | 2.51 | 4.25 |
| Avg len of turn per conv | 21.6 | 71.8 |
| |Voc| of convs | 44134 | 71375 |