convai-challenge/conv_ai_2
收藏数据集概述
数据集描述
数据集摘要
ConvAI 是一个人类与机器人对话的数据集,标记了对话质量。该数据可用于训练评估对话系统的指标。此外,它还可以用于聊天机器人的开发:它包含有关话语和整个对话质量的信息,可以指导对话系统寻找更好的回答。
支持的任务和排行榜
[更多信息需要]
语言
[更多信息需要]
数据集结构
数据实例
json { "dialog_id": "0x648cc5b7", "dialog": [ { "id": 0, "sender": "participant2", "text": "Hi! How is your day? ud83dude09", "sender_class": "Bot" }, { "id": 1, "sender": "participant1", "text": "Hi! Great!", "sender_class": "Human" }, { "id": 2, "sender": "participant2", "text": "I am good thanks for asking are you currently in high school?", "sender_class": "Bot" } ], "bot_profile": [ "my current goal is to run a k.", "when i grow up i want to be a physical therapist.", "im currently in high school.", "i make straight as in school.", "i won homecoming queen this year." ], "user_profile": [ "my favorite color is red.", "i enjoy listening to classical music.", "im a christian.", "i can drive a tractor." ], "eval_score": 4, "profile_match": 1 }
数据字段
dialog_id: 对话的唯一ID。dialog: 对话数组。bot_profile: 用于评估的机器人标注响应。user_profile: 用于评估的用户标注响应。eval_score: (1,2,3,4,5) 用户对对话的喜好程度。缺失值用-1替换。profile_match: (0,1) 用户被提供两个个人描述(每个4句话),其中一个是对话中的机器人给出的,另一个是随机的;用户需要选择其中一个。缺失值用-1替换。
数据分割
[更多信息需要]
数据集创建
策划理由
[更多信息需要]
源数据
初始数据收集和规范化
[更多信息需要]
源语言生产者是谁?
[更多信息需要]
标注
标注过程
[更多信息需要]
标注者是谁?
[更多信息需要]
个人和敏感信息
[更多信息需要]
使用数据的注意事项
数据集的社会影响
[更多信息需要]
偏见的讨论
[更多信息需要]
其他已知限制
[更多信息需要]
附加信息
数据集策展人
[更多信息需要]
许可信息
[更多信息需要]
引用信息
bibtex @article{DBLP:journals/corr/abs-1902-00098, author = {Emily Dinan and Varvara Logacheva and Valentin Malykh and Alexander H. Miller and Kurt Shuster and Jack Urbanek and Douwe Kiela and Arthur Szlam and Iulian Serban and Ryan Lowe and Shrimai Prabhumoye and Alan W. Black and Alexander I. Rudnicky and Jason Williams and Joelle Pineau and Mikhail S. Burtsev and Jason Weston}, title = {The Second Conversational Intelligence Challenge (ConvAI2)}, journal = {CoRR}, volume = {abs/1902.00098}, year = {2019}, url = {http://arxiv.org/abs/1902.00098}, archivePrefix = {arXiv}, eprint = {1902.00098}, timestamp = {Wed, 07 Oct 2020 11:09:41 +0200}, biburl = {https://dblp.org/rec/journals/corr/abs-1902-00098.bib}, bibsource = {dblp computer science bibliography, https://dblp.org} }
贡献
感谢 @rkc007 添加此数据集。




