RuyuanWan/Politeness_Disagreement

---
annotations_creators:
- crowdsourced
language:
- en
language_creators:
- crowdsourced
license: []
multilinguality:
- monolingual
pretty_name: RuyuanWan/Politeness_Disagreement
size_categories: []
source_datasets:
- extended
tags: []
task_categories:
- text-classification
task_ids: []
---

This dataset is processed version of Stanford Politeness Corpus (Wikipedia) including text and the annotation disagreement labels. <br>

Paper: Everyone's Voice Matters: Quantifying Annotation Disagreement Using Demographic Information <br>
Authors: Ruyuan Wan, Jaehyung Kim, Dongyeop Kang <br>
Github repo: https://github.com/minnesotanlp/Quantifying-Annotation-Disagreement <br>

Source Data: [Wikipedia Politeness Corpus(Danescu-Niculescu-Mizil et al. 2013)](https://convokit.cornell.edu/documentation/wiki_politeness.html) <br>

原始信息汇总

数据集概述

基本信息

• 名称: RuyuanWan/Politeness_Disagreement
• 语言: 英语 (en)
• 语言创建者: 众包
• 多语言性: 单语
• 任务类别: 文本分类

数据来源

• 源数据集: 扩展自Wikipedia Politeness Corpus (Danescu-Niculescu-Mizil et al. 2013)

数据处理

• 处理版本: 包含文本及标注分歧标签的斯坦福礼貌语料库处理版本

相关文献

• 论文: Everyones Voice Matters: Quantifying Annotation Disagreement Using Demographic Information
• 作者: Ruyuan Wan, Jaehyung Kim, Dongyeop Kang

许可证

• 许可证信息: 未明确列出

搜集汇总

背景与挑战

背景概述

该数据集是斯坦福礼貌语料库（维基百科）的处理版本，专注于量化注释分歧，包含2,174条英文文本，每条文本都有二元分歧标签（0或1）和分歧率（0到1之间的浮点数）的标注。数据集用于文本分类任务，旨在研究礼貌表达中的分歧现象，并探索人口统计信息对分歧的影响。