ORKG Scientific Synthesis Generation Dataset

该数据集包含了基于开放研究知识图谱（ORKG）中提取的研究问题和相关论文样本，旨在用于科学综合生成的实践。数据集覆盖了多样化的研究问题，其结构设计旨在促进科学综合的生成，同时允许用户输入灵活地反映现实世界中的研究查询。该数据集规模为348个样本，每个样本包含针对一个特定研究问题的五篇论文。所涉及的任务是科学综合生成，这是一种特殊的多文档摘要形式。

This dataset comprises research questions and relevant paper samples extracted from the Open Research Knowledge Graph (ORKG), and is developed for the practice of scientific synthesis generation. The dataset covers a diverse range of research questions, with its structure designed to facilitate the creation of scientific syntheses while allowing flexible user inputs that reflect real-world research queries. The dataset contains a total of 348 samples, where each sample includes five papers corresponding to a specific research question. The core task involved is scientific synthesis generation, a specialized form of multi-document summarization.