Best Practice or Case Studies for Broad-range capture of non-traditional research outputs

Hi All

tl;dr Looking for best practice descriptions or case studies of approaches for capturing and recording (and where necessary minting PIDs) for non traditional research outputs

We are embarking on an effort to capture a PID-centric view of as broad a range as feasible of university outputs. Our basic concept is to build out from available PID-graph resources which cover journal article outputs with a combinatio of encouraging better PID assignment practice, improved capture and curation of outputs in ORCID, and developing simple workflows for minting PIDs for non-traditional research outputs (NTROs) that capture as much consistent metadata as feasible while minimising the burden on researchers and allowing as much flexibility as possible.

We are starting with our own school that has outputs ranging from journal articles, to books, to paintings, exhibitions, theatre performances, media productions, reports and others. It therefore includes physical objects and conceptual objects (performances) where the canonical work is not digital, as well as many cases where it is unclear (eg is the canonical work for a TV news story the video or its broadcast?).

Our straw-person approach is to start with common work types, explore recording these in a PID-providing repository (eg Zenodo) with the goal of providing support and tools to make this as easy as feasible. Minted work-PIDs would then flow into ORCID records providing a structured record and means of collation that is under the control of the researcher to express and surface the works that they feel represent their scholarship.

Using Zenodo as an example there are many options we could adopt for various kinds of work that are plausible choices (object types include videos, physical objects, books etc). What we are interested in is to see existing cases where those choices have been made, any emerging best practice or guidance on how best to represent different object types (particularly where the canonical work is not digital, so GLAM examples will be highly relevant).

Specifically looking for detailed implementations and those that people feel are working well in the real world (or indeed decisions that they would make differently given another chance!). So looking for eg “this is the best way to mint a PID for a contemporary painting, taking account of concerns over copyright and loss of contro over the image in digital form” and not so much “you should mint a PID for everything” or “here’s how to persuade people mint PIDs for everything”

1 Like

Hi Cameron, the NISO Video & Audio Metadata Working Group might be a good starting point for some of this? They’re working on a Recommended Practice that will be shared for public comment (hopefully later this year). And at the NISO Plus conference this year there was also interest in developing standardized metadata for indigenous knowledge - I’ll let you know if anything comes of that. Have also tweeted your request :slight_smile:


I am not sure that I can help with performances, but biologists tend to make things like plasmids, cell lines, antibodies, software tools, instruments, and mice (or other vermin) that other scientists can reuse. These are covered by the RRID initiative and already have good metadata, nearly mapped to DataCite schema, that can be used if you have biologists in your “pile” of people to give non-traditional credit to.

If your sample includes not so many of these kinds of people, then please ignore my note. Otherwise please poke the resolver service to find out what the data schema is for your favorite non-traditional scholarly output such as making a bit of DNA that other people like.


1 Like