基于科学数据的合作网络研究——以ClinicalTrials.gov临床试验数据为例

目的/意义]基于科学数据构建合作网络，并与传统出版物合作网络进行比较，从网络分析层面解读两个合作网络的差异，为科学数据管理工作提供借鉴。方法/过程]以ClinicalTrials.gov网站的临床科学数据库为例，利用爬虫抓取该网站上传统论文题录信息以及临床试验信息的元数据并分别构建合作网络，通过复杂网络分析比较试验合作机构网络与论文合作机构网络之间的异同。结果/结论]基于科学数据集和论文数据集的元数据构建的合作网络，与仅从论文数据集中提取元数据构建的网络相比，前者能够展现更丰富准确的合作信息，从而揭示科学数据管理和开放共享的重要性。

Research on Collaboration Network of Scientific Data-Taking Clinical Trial Data of ClinicalTrials.gov as an Example

Purpose/significance] Based on the scientific data, we constructed a new collaboration network and compared with cooperative network of traditional publications. By exploring the differences of the two collaboration network from the perspective of network analysis, this paper provided reference for scientific data management. Method/process] Taking clinical science database of ClinicalTrials.gov website as an example, this paper used crawler technology to capture the metadata of traditional publication and clinical trial research. Then, based on these two kinds of metadata, this paper constructed different cooperative networks respectively. Finally, it used complex network analysis to explore these networks to compare the similarities and differences between two networks. Result/conclusion] Scientific collaboration network extracted metadata from a scientific repository and publication could provide richer and more accurate information of collaboration than just using metadata from publications alone.This paper reveals the importance of scientific data management and open sharing.