The following proposal was approved by the MaRDA Council on 2021-04-15. The effective start date is 2021-05-01.

• Donny Winston, Polyneme LLC
• June Lau, NIST
• Ken Kroenlein, Citrine Informatics

Shared datasets often have column/field names that are ambiguous in their meaning, or contain identical/related concepts with different names, hindering reuse. For example in electron microscopy, "high-tension" and "acceleration voltage" are used interchangeably (identical concepts, different names). However, "magnification" and "field-of-view" are not the same thing, but are closely related (domain experts know that dividing the field-of-view by pixel size will return the magnification value). This ambiguity happens regardless of the method of sharing -- via files, web pages, or APIs. The traditional solution for this is to provide documentation.

However, documentation is typically delivered “out-of-band”, i.e., not directly linked within the same artifact (file/object) as the data. Thus, a data consumer needs to link terms to their definitions by eye, i.e., one reads many pages of documentation based on one’s needs, creates an ontological map in one’s mind that must be re-created again and again (which should best be replaced by a machine accessible ontological map) - which raises the barrier to data reusability. Furthermore, any linking among terms by different data providers, whether to assert equivalence or to clarify differences, is typically done in an ad hoc manner, hindering interoperability and thus reuse.

This working group will produce a Working Group Note¹ for sharing materials research data as 5-Star Linked Data. Specifically, the working group will:

1. run a pilot study using datasets from at least three (3) independent materials research data providers. The working group will split into interdependent groups, one per dataset, to ensure:

   • a data dictionary is created and deployed online, with resolvable HTTP URLs for each term, that documents the fields of the dataset;
   • the dictionary for each dataset includes qualified links to terms of another dataset’s dictionary. In the cases where terms are related but not equivalent, those relationships are explained and documented; and
   • the dataset itself is made accessible as Linked Data, and thus self-describing in linking to its data dictionary.

2. produce a Working Group Note that uses the completed pilot study as a case study in order to document learnings in (a) creating and deploying a data dictionary for a materials research dataset; (b) relating a dictionary’s terms to external terms in order to provide richer context for a data consumer; and (c) producing a self-describing distribution of a dataset by linking to its data dictionary.

Proposed timeline (12 months): Six months for Activity 1 (Pilot study), with two months allocated per sub-activity (isolated data dictionaries, linked data dictionaries, and linked data). A further six months to produce the Working Group Note (rounds of Working Drafts, iterating on pilot study activities, etc.).

• Produce practical recommendations for how to deliver dataset documentation in-band, i.e., metadata that travels with the data.
• Produce practical recommendations for how to enrich a dataset’s context by making relationships to other datasets in the materials research community explicit.
• The expected impact is to begin to improve materials research dataset interoperability and reuse (c.f. FAIR) by identifying a practical method whereby data dictionaries can be published, inter-linked, and referenced by the datasets they describe. Such a bottom-up “lightweight ontology” approach should complement existing top-down “upper ontology” efforts.

• A Working Group Note, published on the MaRDA Alliance website,12 months after commencement of the working group, as outlined in the Work Plan.
• Recorded webinars at approx. 6 and 12 months – the latter linked to the published Note, wherein the working group presents to MaRDA members in order to answer questions and encourage wider adoption of the group’s recommendations by the community.

There are currently ten (10) other working group members. They are not listed here to respect privacy. If you are a working group member and wish to have your name listed here, let Donny know.

1: Terminology references the W3C Process. While the produced Working Group Note may later advance in maturity level towards e.g. a Candidate Recommendation, the scope of this working group is to produce a demonstrative technical report.