pchristm/CompMix
收藏资源简介:
--- license: cc-by-4.0 task_categories: - question-answering - conversational language: - en tags: - complex - question answering - complexQA - QA - heterogeneous sources pretty_name: CompMix size_categories: - 1K<n<10K splits: - name: train num_examples: 4966 - name: validation num_examples: 1680 - name: test num_examples: 2764 --- # Dataset Card for ConvMix ## Dataset Description - **Homepage:** [CompMix Website](https://qa.mpi-inf.mpg.de/compmix) - **Point of Contact:** [Philipp Christmann](mailto:pchristm@mpi-inf.mpg.de) ### Dataset Summary CompMix collates the completed versions of the conversational questions in the [ConvMix dataset](https://convinse.mpi-inf.mpg.de), that are provided directly by crowdworkers from Amazon Mechanical Turk (AMT). Questions in CompMix exhibit complex phenomena like the presence of multiple entities, relations, temporal conditions, comparisons, aggregations, and more. It is aimed at evaluating QA methods that operate over a mixture of heterogeneous input sources (KB, text, tables, infoboxes). The dataset has 9,410 questions, split into train (4,966 questions), dev (1,680), and test (2,764) sets. All answers provided in the CompMix dataset are grounded to the KB (except for dates which are normalized, and other literals like names). Further details will be provided in a dedicated write-up soon. ### Dataset Creation CompMix collates the completed versions of the conversational questions in ConvMix, that are provided directly by the crowdworkers. The ConvMix benchmark, on which CompMix is based, was created by real humans. We tried to ensure that the collected data is as natural as possible. Master crowdworkers on Amazon Mechanical Turk (AMT) selected an entity of interest in a specific domain, and then started issuing conversational questions on this entity, potentially drifting to other topics of interest throughout the course of the conversation. By letting users choose the entities themselves, we aimed to ensure that they are more interested into the topics the conversations are based on. After writing a question, users were asked to find the answer in eithers Wikidata, Wikipedia text, a Wikipedia table or a Wikipedia infobox, whatever they find more natural for the specific question at hand. Since Wikidata requires some basic understanding of knowledge bases, we provided video guidelines that illustrated how Wikidata can be used for detecting answers, following an example conversation. For each conversational question, that might be incomplete, the crowdworker provides a completed question that is intent-explicit, and can be answered without the conversational context. These questions constitute the CompMix dataset. We provide also the answer source the user found the answer in and question entities.
--- license: CC BY 4.0 task_categories: - 问答(question-answering) - 会话式(conversational) language: - 英语(en) tags: - 复杂(complex) - 问答(question answering) - 复杂问答(complexQA) - QA(QA) - 异构数据源(heterogeneous sources) pretty_name: CompMix size_categories: - 1000 < 样本数 < 10000 splits: - name: train num_examples: 4966 - name: validation num_examples: 1680 - name: test num_examples: 2764 --- # 《CompMix数据集卡片》 ## 数据集说明 - **主页**:[CompMix官方网站](https://qa.mpi-inf.mpg.de/compmix) - **联系人**:[Philipp Christmann](mailto:pchristm@mpi-inf.mpg.de) ### 数据集概述 CompMix整合了[ConvMix数据集](https://convinse.mpi-inf.mpg.de)中会话式问题的完整版本,这些完整问题由亚马逊机械Turk(Amazon Mechanical Turk,简称AMT)平台的众包工作者直接提供。CompMix中的问题包含多种复杂现象,例如多实体、多关系、时序条件、比较、聚合等。该数据集旨在评估可处理混合异构输入源(知识库(KB)、文本、表格、信息框)的问答方法。数据集共包含9410个问题,划分为训练集(4966个问题)、验证集(1680个问题)与测试集(2764个问题)。CompMix数据集中的所有答案均锚定至知识库(KB)(日期已做归一化处理,姓名等其他字面量除外)。 更多细节将在后续的专门研究论文中公布。 ### 数据集构建 CompMix整合了ConvMix数据集中会话式问题的完整版本,这些版本由众包工作者直接提供。 CompMix所基于的ConvMix基准数据集由真实人类参与者构建。我们力求确保所收集的数据尽可能自然。亚马逊机械Turk(AMT)平台的资深众包工作者会在特定领域中选取一个感兴趣的实体,随后围绕该实体发起会话式问题,在对话过程中也可能转向其他感兴趣的话题。通过让用户自主选择实体,我们旨在确保参与者对对话所围绕的话题拥有更高的兴趣度。编写完问题后,参与者需要从维基数据(Wikidata)、维基百科文本、维基百科表格或维基百科信息框中寻找对应问题的答案,选择他们认为对当前问题最自然的来源即可。由于使用维基数据需要具备一定的知识库基础知识,我们提供了视频指南,通过示例对话演示如何利用维基数据查找问题答案。针对每个可能不完整的会话式问题,众包工作者需要提供一个意图明确、无需依赖会话上下文即可回答的完整问题,这些完整问题便构成了CompMix数据集。我们同时提供了参与者查找答案所用的来源以及问题所涉及的实体。
数据集概述
基本信息
- 许可证: cc-by-4.0
- 任务类别:
- 问答
- 对话
- 语言: 英语
- 标签:
- 复杂
- 问答
- complexQA
- QA
- 异构来源
- 美观名称: CompMix
- 数据集大小: 1K<n<10K
数据集结构
- 训练集: 4,966个问题
- 验证集: 1,680个问题
- 测试集: 2,764个问题
数据集内容
- 数据来源: 由Amazon Mechanical Turk的众包工作者直接提供
- 问题特点: 包含多实体、关系、时间条件、比较、聚合等复杂现象
- 答案基础: 所有答案均基于知识库(KB),日期和其他文字(如名称)除外
数据集创建
- 创建方式: 基于ConvMix基准,由众包工作者选择感兴趣的实体,并围绕该实体发起对话式问题
- 答案查找: 用户根据问题在Wikidata、Wikipedia文本、表格或信息框中查找答案
- 额外提供: 提供用户找到答案的来源和问题实体




