| title | layout | filename |
|---|---|---|
Self-assessment interface using FAIRMetrics |
template |
interop-wg/linked-data-webapp/fm_demo |
The FAIR Evaluator framework [Wilkinson, Dumontier et al., Scientific Data 6:174] is composed of Maturity Indicators (MI), compliance tests and the evaluator application itself. The authors expect growing contributions in MI specifications and compliance tests to improve relevance and coverage of FAIR evaluation. They also propose that these tests should be carried out by third party organisations. In spite of many research contributions highlighting the potential benefits of FAIR principles for the research community and their potential impacts, few efforts have been done so far to take advantage from their concrete implementation, in the process of improving FAIRness of users/community resources. Early prototypes focus on publishing FAIRness metrics computed offline for predefined sets of public web accessible resources. However this does not provide concrete help to developers for better sharing their published works. In this work we propose a web demonstrator, leveraging existing web APIs, aimed at i) evaluating FAIR maturity indicators and ii) providing hints to progress in the FAIRification process.
For now, we are willing to use the FAIRMetrics smartapi to create a simple interface and provide the user with an easy way to verify how FAIR is his resource. We want to provide the user with clear informations on how to validate each aspect of the FAIR principles and help him to improve the FAIRness of his resource. We address the following use-case. A researcher/developer just produced new results in terms of data, software, or web-accessible documents. She wants to assess, for instance, how findable or reusable is the proposed resource. She submits to the self-assessment tool the resource identifier or web location (URL). This tool dynamically computes the metrics as specified by the community initiatives and provides the users with a score for each metrics. A particular attention is paid to provide users with explanation on why a particular metrics is failing, and which technical solution could be adopted to enhance the FAIR metrics evaluation. We propose that the results of such FAIR metrics evaluations remain private to better help resource creators and developer in their repeated efforts towards more findable and reusable machine-actionable resources.
Our tool is available as a web demonstrator, hosted by the Elixir-FR node. The tool has been developed in Python with the Flask framework. It leverages the REST protocol to send queries to the FAIR evaluation framework APIs resulting from community efforts. The outputs of each MI metric evaluation is parsed and presented in a user-friendly way. Links to useful resources for improving the resource FAIRness are provided.
In this work we propose a simple web tool, aimed at supporting resource creators and developers in their sharing effort to better conform to standardization efforts such as the FAIR principles. s future works we plan to i) investigate how some of the metrics could be more efficiently computed and ii) join already started community efforts towards transparent and explainable automated computation of FAIR metrics.
The aim of this demonstration is to provide a typical usecase for this project idea.
A researcher created a fresh new dataset and host it on the web, associated to a URL or DOI.
The researcher want to have an idea of how FAIR is his resource, to maybe improve it further.
The researcher submit his resource throught a URL or DOI with minimal informations required (no personnal informations).
Provide the researcher with a one page result, with negative or positive score for each Maturity Indicator.
When the result is negative, explain where it failed, and provide suggestions of how to validate this Maturity Indicator. This will provide a usefull help to the researcher, and then allow him to improve the FAIRness of his ressource.
The reasercher will try to improve the FAIRness of his ressource, and test it again, and again, trying to validate the MI(s) he is interested in.