Improving Impact Metrics of Open and Free Biodiversity Data through Linked Metadata and Academic Outreach
Improving Impact Metrics of Open and Free Biodiversity Data through Linked Metadata and Academic Outreach
复制标题
通过链接元数据和学术推广提高开放和免费生物多样性数据的影响力指标
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Daniel Noesgaard
中科院分区：
文献类型：
--
作者：
Daniel Noesgaard
收集、清理和发布生物多样性数据集所需的工作非常重要，那些从事这些工作的人的努力值得得到认可。研究人员利用全球生物多样性信息设施 GBIF 提供的开放生物多样性数据发表研究成果，速度约为每天两篇论文。这些研究涵盖宏观生态学、进化论、气候变化和外来入侵物种等领域，依赖于数百家出版机构的数据共享和数千名个人贡献者的策展工作。其中 90% 以上的数据集均获得知识共享归属许可（CC BY 和 CC BY-NC）许可，数据用户必须注明数据集提供商的身份。对于 GBIF 来说，至关重要的是，将这些科学用途与基础数据联系起来，作为展示开放科学的价值和影响的一种手段，同时寻求确保个人、组织和国家对全球生物多样性开放数据池的贡献。
从 GBIF.org 中下载的每一次经过身份验证的事件记录都会分配一个唯一的数字对象标识符 (DOI)。这些 DOI 每个解析为包含详细信息的登录页面
用于生成下载的搜索参数
有助于下载的基础数据集的定量图
包含在依赖数据的作品中的简单引用
用于生成下载的搜索参数
有助于下载的基础数据集的定量图
包含在依赖数据的作品中的简单引用
当作者正确使用并由期刊正确存储在文章元数据中时，DOI 引文会在科学论文和基础数据之间建立直接链接。 Crossref（学术文献的主要 DOI 注册机构）在事件数据中公开此类链接，可以通过编程方式使用这些链接来报告单个数据集的直接使用情况。 GBIF 还记录这些链接，永久保留下载档案，同时在下载登陆页面上公开引用计数，该计数也在每个贡献数据集和出版商的登陆页面上进行汇总。可以扩展引用计数以生成与特定数据集的使用明确链接的所有论文的列表。
2018 年，只有 15% 基于 GBIF 介导的数据的论文使用 DOI 来引用或承认研究中使用的数据集。为了促进数据出版商的信用和数据共享的数字认可，自 2018 年 4 月以来，只要论文未能包含正确的数据引用，GBIF 秘书处就会系统地与作者和出版商联系。虽然出版滞后可能会阻碍立竿见影的效果，但初步调查结果表明，采用率正在提高——2019 年上半年，DOI 数据引用的论文数量比 2018 年增加了 60% 以上。
本演示将重点关注链接科学出版物和数据的价值，探讨通过 DOI 元数据建立自动链接的潜力，同时展示通过向作者和期刊出版商开展推广活动来改进数据使用指标和数据提供者归属的努力。
The work required to collect, clean and publish biodiversity datasets is significant, and those who do it deserve recognition for their efforts. Researchers publish studies using open biodiversity data available from GBIF—the Global Biodiversity Information Facility—at a rate of about two papers a day. These studies cover areas such as macroecology, evolution, climate change, and invasive alien species, relying on data sharing by hundreds of publishing institutions and the curatorial work of thousands of individual contributors. With more than 90 per cent of these datasets licensed under Creative Commons Attribution licenses (CC BY and CC BY-NC), data users are required to credit the dataset providers. For GBIF, it is crucial to link these scientific uses to the underlying data as one means of demonstrating the value and impact of open science, while seeking to ensure attribution of individual, organizational and national contributions to the global pool of open data about biodiversity.
Every single authenticated download of occurrence records from GBIF.org is issued a unique Digital Object Identifier (DOI). These DOIs each resolve to a landing page that contains details of
the search parameters used to generate the download
a quantitative map of the underlying datasets that contributed to the download
a simple citation to be included in works that rely on the data
the search parameters used to generate the download
a quantitative map of the underlying datasets that contributed to the download
a simple citation to be included in works that rely on the data
When used properly by authors and deposited correctly by journals in the article metadata, the DOI citation establishes a direct link between a scientific paper and the underlying data. Crossref—the main DOI Registration Agency for academic literature— exposes such links in Event Data, which can be consumed programmatically to report direct use of individual datasets. GBIF also records these links, permanently preserving the download archives while exposing a citation count on download landing pages that is also summarized on the landing pages of each contributing datasets and publishers. The citation counts can be expanded to produce lists of all papers unambiguously linked to use of specific datasets.
In 2018, just 15 per cent of papers based on GBIF-mediated data used DOIs to cite or acknowledge the datasets used in the studies. To promote crediting of data publishers and digital recognition of data sharing, the GBIF Secretariat has been reaching out systematically to authors and publishers since April 2018 whenever a paper fails to include a proper data citation. While publishing lags may hinder immediate effects, preliminary findings suggest that uptake is improving—as the number of papers with DOI data citations during the first part of 2019 is up more than 60 per cent compared to 2018.
Focusing on the value of linking scientific publications and data, this presentation will explore the potential for establishing automatic linkage through DOI metadata while demonstrating efforts to improve metrics of data use and attribution of data providers through outreach campaigns to authors and journal publishers.