Your Name and Title: Mr Samuel Simango (Senior Manager: Academic Support Services)
Library, School, or Organization Name: Rhodes University
Co-Presenter Name(s): N/A
Area of the World from Which You Will Present: South Africa
Language in Which You Will Present: English
Target Audience(s): Students, researchers, data stewards, data curators, RDM librarians
Short Session Description (one line): This session will demonstrate how generative AI was used to translate complex Science Europe policy into an intuitive, multiple-choice Data Management Plan template deployed within the FAIR-compliant FAIRWizard software platform
Full Session Description (as long as you would like):
This presentation relates to the application of generative Artificial Intelligence to the development of a bespoke user-friendly Data Management Planning template which complies the FAIR Data Principles. A data management plan is an a priori description of how data will be handled both during and after research. While Data Management Plans (DMPs) are critical for ensuring research data is Findable, Accessible, Interoperable, and Re-usable (FAIR), they have traditionally made use of templates which focused on addressing pure Research Data Management issues rather than FAIRness. Furthermore, such DMPs have relied on the use of templates which have not necessarily been user-friendly in nature.
In the recent past, Science Europe has taken steps to simplify the creation of DMPs by developing a standardised practical guide focused on Data Management Planning. This guide contains a framework for the standardisation of DMP templates known as the Core Requirements for Data Management Plans (the Core Requirements). These requirements set out the structure and core category of questions which can be answered in order to create Data Management Plans. The requirements are supported by a set of DMP evaluation rubrics which provide guidance to researchers and reviewers of DMPs.
A critical strength of the Core Requirements is their ability to be mapped directly to the different elements of the FAIR Data Principles – an attribute which creates alignment between Research Data Management elements on the one hand and the FAIR Data Principles on the other hand. This attribute can help researchers create DMPs which are FAIR compliant if they simply make use of a DMP template which is based on the Core Requirements. However, this task is plagued by two problems. Firstly, most DMP software platforms were not designed from the ground up to be FAIR compliant. Secondly, in instances where such tools do exist, they do not contain a built-in standard template which is based on the Core Requirements.
This project attempted to address these two issues by designing a user-friendly DMP template, which is based on the Core Requirements, and then deploying the template by using FAIR Wizard – a DMP software product which automatically enables the assessment of compliance with the FAIR Data Principles. In the process, reliance was placed on Gemini in order to convert the Core Requirements and its associated reviewers’ DMP evaluation rubric into a set of multiple-choice questions. The end result is a potentially more accessible user-friendly Data Management Planning template which complies the both the Core Requirements for DMPs as well as the FAIR Data Principles. Such a template can be used by the creators of DMPs (i.e., students and researchers) as well as the reviewers of the DMPs (i.e., funders, academic supervisors, members of research ethics committees, data stewards, data curators, RDM librarians). The presentation will synthesise the nature of the problem, describe the manner in which the problem was addressed and also provide a practical demonstration of end-result – the AI-designed template within FAIR Wizard.
Keywords: Generative Artificial Intelligence, Research Data Management, Data Management Plans, FAIR Data Principles, Science Europe Core Requirements for DMPs
Websites / URLs Associated with Your Session: https://rul.fair-wizard.com/wizard/projects/b04e39db-c695-455f-9a69-4b14ffe84f29
Replies