WebQuestionsSP

WebQuestionsSP 数据集作为我们 ACL-2016 论文“知识库问答的语义分析标签的价值”[Yih, Richardson, Meek, Chang & Suh, 2016] 的一部分发布，其中我们评估了收集语义分析的价值, 与答案，针对最初来自 WebQuestions [Berant et al., 2013] 的一组问题。 WebQuestionsSP 数据集包含 SPARQL 查询中 4,737 个问题的完整语义解析，以及剩余 1,073 个问题的“部分”注释，这些问题无法进行有效解析，或者问题本身不好或需要描述性答案。此版本还包括评估脚本和 STAGG 语义解析系统在使用完整语义解析进行训练时的输出。更多详细信息可在此版本中包含的文档和标签说明以及论文中找到。

WebQuestionsSP dataset was released as part of our ACL 2016 paper "The Value of Semantic Parse Labels for Knowledge Base Question Answering" [Yih, Richardson, Meek, Chang & Suh, 2016], where we evaluated the value of collecting semantic parse labels compared to answer-only annotations for a set of questions originally sourced from WebQuestions [Berant et al., 2013]. The WebQuestionsSP dataset contains full semantic parses for 4,737 questions formulated as SPARQL queries, alongside "partial" annotations for the remaining 1,073 questions, which either cannot be effectively parsed, are poorly formulated, or require descriptive answers. This release also includes evaluation scripts and the outputs of the STAGG semantic parsing system when trained on full semantic parses. Further details can be found in the documentation and label descriptions included with this release, as well as in the aforementioned paper.

搜集汇总

背景与挑战

背景概述

WebQuestionsSP是一个用于知识库问答的数据集，包含4,737个问题的完整语义解析（SPARQL查询）和1,073个问题的部分注释，适用于评估语义分析在问答系统中的价值。该数据集源自ACL-2016论文，附带评估工具和预训练系统输出，支持问答和文本分类任务。