Agile Research Data Management with FDOs using LinkAhead

Author:

Fitschen Timm,tom Wörden Henrik,Schlemmer Alexander,Spreckelsen FlorianORCID,Hornung Daniel

Abstract

One essential question with regard to the implementation of FAIR (Wilkinson et al. 2016) Digital Objects (FDOs) in everyday research is the following: How is data that is acquired in some way transformed into FDOs? Creating FDOs from data is a two-fold problem: "FAIR principles are policies, whereas the digital objects are technical abstractions" (Schwardmann 2020). Regarding the technical side, in order to become FDOs, raw data stored in files and databases have to be bundled with their metadata and PIDs have to be assigned. With good tools at hand, sharing data as an FDO with others might only be a matter of a few mouse clicks -- if the metadata is readily available. However, the process of collecting metadata comes with significant challenges of its own. While sometimes necessary, the manual annotation with metadata is error-prone and time-consuming. Due to resource constraints and time pressure, researchers might skip this task whenever it does not have any direct benefits for their work in the time frame of their current project. The experience with existing data repositories tells us that adding metadata at a late stage in the research data life cycle (for example just before publication) delays the problem in the best case. In the worst case, important information has already been lost at that stage. Furthermore, there are the FAIR principles which, for researchers, mean more rules to follow and thus more time spent on data management. Research data should be enriched with FAIR metadata as early as possible to ensure that the research data is FDO-ready when needed. In order to do this, researchers need tools that assist them with the task of making their data FDO-ready and those tools must not hinder the research process but in the best case even promote it. This means that the drawbacks of making data FDO-ready need to be mitigated and compensated by direct benefits to researchers. In this contribution, we present how early-on FDO-readyness can be achieved with the open source research data management toolkit LinkAhead and how researchers profit from the FDO-readyness directly in their work. LinkAhead, a CaosDB (Fitschen et al. 2019) distribution, assists its users from the very first steps of data acquisition to the completion of FDOs and data publication by means of a semantic data modal, metadata annotation to raw data and a powerful search capabilities. Why would researchers do what is necessary to make their data FOD-ready, early-on? With LinkAhead, the FDO-readyness is a welcome side-effect for users. Even though LinkAhead cannot magically generate all relevant metadata and make data FAIR, LinkAheads allows the automation of the process where possible and assists users elsewhere. The inevitable additional work for researchers is reduced as well as compensated with new possiblities for users to work with their data. Thus users are nudged into storing their data in clear and understandable structures and into annotating their data with high-quality metadata. We will highlight in the following, how users benefit from early FDO-readiness in their daily work due to those characteristics of LinkAhead. LinkAhead adapts to the changing needs of the researchers. It thus allows research data management to be an agile process and ensures that researchers can efficiently conduct their daily work. At the same time, it supports the development, documentation and observance of standards which is vital for the commensurability, reusability, and reproducibility of research findings. LinkAhead is designed to be the first tool after data acquisition and the last tool before the publication of data. It can be fed with data from LIMS, ELNs, simulation and analysis software, helps with automation of workflows, and manages raw data in files. Which direct benefit can LinkAhead offer to its users if they do what is necessary to make FDOs from data? When searching for data in general or FDOs in particular researchers can employ metadata and the connections among data in order to find what they are looking for. Thereby, browsing the data for example in the LinkAhead web interface can be very targeted. Additionally, these search capabilities can be used within analysis workflows in order to create the correct basis of data using FDOs directly for the question at hand. Client libraries, like the Python client, allow to include this into automated analyses. Since manual data insertion is inefficient in many research environments, LinkAhead does not only offer the insertion of data via web forms, but encourages the usage of automatic processes like the LinkAhead Crawler. While metadata should be added already during this insertion step if possible, LinkAhead assists in completing metadata after the initial insertion in order to strike a balance between interrupting the research workflow and running into the above mentioned challenges when adding metadata too late. This automatic data insertion process is highly customizable and allows to complement data as soon as possible such that FDOs are constituted. The semantic data model of LinkAhead allows researchers to use ontologies of their domain but also to extend those where necessary for the work at hand. This allows an agile adaption to changed requirements or new challenges and LinkAhead assures compatibility with old data if possible. The data model can capture both relations within an FDO, among metadata, data, and possibly files and references to other FDOs directly within LinkAhead or FDOs stored elsewhere using PIDs. The semantic data model and additional constraints facilitate the creation and validation of FDOs. LinkAhead allows to seamlessly integrate the FDO concept into the workflows of LinkAhead users. Thereby collecting information necessary for FDOs is not a burden for the researcher but the information can be directly used. The search capabilities of LinkAhead can employ the metadata of FDOs and the connections to other FDOs and their metadata. Components of the LinkAhead toolkit like the web interface allow users to access FDOs in an intuitive way while the LinkAhead API allows the direct use of data (and FDOs) in automatic processing and analysis. Thus, LinkAhead is a tool which does not only assist in the technical process of creating digital objects, it also creates incentives for its target users to adhere to FAIR guiding principles. It brings the benefits of FDOs to the people who have to do the extra work. Strategically, this is of utmost importance if the FDO initiative as a whole is to succeed.

Publisher

Pensoft Publishers

Subject

General Medicine

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3