人工智能培训

搜索

论文代码开源:基于深度生成建模的可扩展种群综合(Scalable Population Synthesis with Deep Generative Model

[复制链接]
admin 发表于 2018-8-25 09:25:32 | 显示全部楼层 |阅读模式
admin 2018-8-25 09:25:32 427 0 显示全部楼层
人工智能论文代码开源:基于深度生成建模的可扩展种群综合(Scalable Population Synthesis with Deep Generative Modeling)请注意该人工智能论文代码开源在github,大部分是python写的,框架可能是tensorflow或者pytorch。人类通过参与涉及一系列相互关联的问题和答案的对话来收集信息。对于协助信息收集的机器,因此必须使他们能够回答会话问题。我们介绍了CoQA,这是一个用于构建ConversationalQuestion Answering系统的新数据集。我们的数据集包含127k个问题及答案,这些问题来自7个不同领域的8k文本段落对话。问题是对话的,答案是自由格式文本,并在段落中突出显示相应的证据。我们深入分析了CoQA,并且表明会话问题具有挑战性现象,而不存在不存在的阅读理解数据集,例如共指和实用推理。我们在CoQA中评估强大的会话和阅读理解模型。最好的系统获得了65.1%的F1分数,这是人类表现(88.8%)的23.7分,表明有足够的改善空间。我们推出CoQA作为对社区的挑战
Humans gather information by engaging in conversations involving a series ofinterconnected questions and answers.For machines to assist in informationgathering, it is therefore essential to enable them to answer conversationalquestions.We introduce CoQA, a novel dataset for building ConversationalQuestion Answering systems.Our dataset contains 127k questions with answers,obtained from 8k conversations about text passages from seven diverse domains.The questions are conversational, and the answers are free-form text with theircorresponding evidence highlighted in the passage.We analyze CoQA in depth andshow that conversational questions have challenging phenomena not present inexisting reading comprehension datasets, e.g., coreference and pragmaticreasoning.We evaluate strong conversational and reading comprehension modelson CoQA.The best system obtains an F1 score of 65.1%, which is 23.7 pointsbehind human performance (88.8%), indicating there is ample room forimprovement.We launch CoQA as a challenge to the community atthis http URL论文代码开源:基于深度生成建模的可扩展种群综合(Scalable Population Synthesis with Deep Generative Modeling) uyMC7h7y96cBmzyk.jpg
URL地址:https://arxiv.org/abs/1808.07042v1     ----pdf下载地址:https://arxiv.org/pdf/1808.07042v1    ----         ----github下载地址:https://github.com/chiahsuan156/Question-Answering_Resources_Papers    ----    论文代码开源:基于深度生成建模的可扩展种群综合(Scalable Population Synthesis with Deep Generative Modeling)请注意该人工智能论文代码开源在github,大部分是python写的,框架可能是tensorflow或者pytorch,keras,至于具体是哪一个没有完全测试。
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则 返回列表 发新帖

admin当前离线
管理员

查看:427 | 回复:0

快速回复 返回顶部 返回列表