sled-umich/Conversation-Entailment

Conversation-Entailment数据集旨在解决传统文本蕴含方法在处理对话脚本时的局限性。该数据集包含了对话脚本中的独特特征，如参与者之间的轮换、参与者之间的基础、话语的不同语言现象以及对话含义。数据集的设计是为了支持从对话脚本中自动推断蕴含关系，填补了传统方法在这一领域的不足。

The Conversation-Entailment dataset is developed to address the limitations of traditional textual entailment methods when processing dialogue scripts. This dataset captures unique features inherent in dialogue scripts, including turn-taking between participants, common ground among participants, diverse linguistic phenomena in utterances, and dialogue-level meaning. Designed to enable automatic entailment relation inference from dialogue scripts, the dataset fills the research gap left by traditional approaches in this domain.

原始信息汇总

数据集概述

• 名称: Conversation-Entailment
• 语言: 英语 (en)
• 语言创建者: 众包 (crowdsourced)
• 许可: 未指定
• 多语言性: 单语
• 大小: 小于1千条记录 (n<1K)
• 来源: 原始数据
• 标签:
  • 对话式
  • 蕴含
• 任务类别:
  • 对话式
  • 文本分类

数据集描述

Conversation-Entailment 数据集专注于对话式文本蕴含，旨在解决传统文本蕴含方法在处理对话行为时的局限性。该数据集包含对话脚本，涉及对话参与者之间的轮流、基础、不同语言现象和对话含义。

数据集样本

json { "id": 3, "type": "fact", "dialog_num_list": [30, 31], "dialog_speaker_list": ["B", "A"], "dialog_text_list": ["Have you seen SLEEPING WITH THE ENEMY?", "No. Ive heard, Ive heard thats really great, though."], "h": "SpeakerA and SpeakerB have seen SLEEPING WITH THE ENEMY", "entailment": false, "dialog_source": "SW2010" }

引用信息

tex @inproceedings{zhang-chai-2010-towards, title = "Towards Conversation Entailment: An Empirical Investigation", author = "Zhang, Chen and Chai, Joyce", booktitle = "Proceedings of the 2010 Conference on Empirical Methods in Natural Language Processing", month = oct, year = "2010", address = "Cambridge, MA", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/D10-1074", pages = "756--766", }