Skip to content

STAR

DialogueEnglish

STAR is a dialogue-focused dataset in English that provides 5,82 labeled examples distributed in JSON format.

About STAR

A schema-guided task oriented dialog dataset consisting of 127,833 utterances and knowledge base queries across 5,820 task-oriented dialogs in 13 domains that is especially designed to facilitate task and domain transfer learning in task-oriented dialog.

Details

Task
Dialogue
Language
English
Format
JSON
Rows / instances
5,82
Creator
Mosig et al.
Year
2020
Download Paper

Related Dialogue datasets

FAQ