---
language: ja
license: cc-by-sa-4.0
---

# Abstract

This dataset comprises exchanges of Yes/No questions from the 
system and corresponding user responses, extracted from the multimodal 
dialogue corpus Hazumi.
It was used to fine-tune our [released model](https://huggingface.co/ouktlab/Hazumi-AffNeg-Classifier).