SurveySum
收藏资源简介:
SurveySum数据集由数字经济研究所创建,旨在解决特定领域内多文档摘要工具的缺乏问题。该数据集包含79个部分,每个部分关联平均7.38篇科学文章,通过人工和自动方法从现有调查中提取和验证。数据集的创建过程涉及从科学文献中选择调查,提取文本和引用,并通过API和爬虫获取全文。SurveySum数据集主要应用于科学调查文本的自动生成,旨在提高科学文献处理的效率和准确性。
The SurveySum dataset was developed by the Institute of Digital Economy, aiming to address the shortage of multi-document summarization tools in specific domains. This dataset comprises 79 sections, with an average of 7.38 scientific articles associated with each section. It is extracted and validated from existing surveys through both manual and automated methods. The dataset creation process involves selecting surveys from scientific literature, extracting text and citations, and acquiring full texts via APIs and web crawlers. The SurveySum dataset is primarily applied to the automatic generation of scientific survey texts, with the goal of improving the efficiency and accuracy of scientific literature processing.

- 1SurveySum: A Dataset for Summarizing Multiple Scientific Articles into a Survey Section数字经济研究所,巴西利亚-DF,巴西 · 2024年



