» Datasets

This page is still being developed and, thus, subject to change.
If you want to contribute to the SIGdial resource portal, please contact us at webadmin@sigdial.org.

The table is both searchable and sortable: to search the table, type in a search term and select the field of interest; to sort the table, click a column header of interest and click again to reverse the sorting order.

Name Type Topics Avg. # of turns Total # of dialogues Total # of words Description Links
DSTC1 Spoken Bus schedules 13.56 15000 3700000 Bus ride information system
Let's Go! Spoken Bus schedules 171128 Bus ride information system
DSTC2 Spoken Restaurants 7.88 3000 432000 Restaurant booking system
Acknowledgement

This overview has been adapted from A Survey of Available Corpora for Building Data-Driven Dialogue Systems, with permission; see the survey website for reference and please cite the paper if useful.