Add ResearchOrganization subclass of Organization

[timbl]

We are using the set of subclasses of Organization to allow users to specify items in their Curriculum Vitae.
The subclasses suggested for Organization are

# Airline - a Corporation
# Consortium - a Corporation or a NGO
 schema:Corporation
 schema:EducationalOrganization
# FundingScheme - eh?
 schema:GovernmentOrganization
# LibrarySystem
# LocalBusiness - Corporation
# MedicalOrganization - a Corporation or a NGO
 schema:NGO
 # NewsMediaOrganization - a Corporation or a NGO
schema:PerformingGroup # a band
schema:Project # like Solid
schema:SportsOrganization # a Team
 ) .

where the ones commented out we have suppressed for the moment. The ones
which are live are implemented by mapping them to Wikidata classes, and then using the
WD SPARQL server to drive an autocomplete entry field.

The most important missing subclass is ResearchOrganization.

WD has one https://www.wikidata.org/wiki/Q31855 called Research Institute.
Propose that they be considered equivalent classes.

[danbri]

Thanks @timbl

I've coded this up for our upcoming release, in the "pending" section - which gives us room to refine the definition over time.

As proposed here (and if identical to the Wikipedia type) it is for rather formal research organizations (although Wikidata suggests "environment" as a softer synonym). I am comfortable with this, in part because we also have a softer / vaguer structure under Project, https://schema.org/ResearchProject.

So for a group of internet collaborators figuring something out together in IRC / wiki / github, it might not be a Scientific Institute, but it could still be a Research Project. It would be good to explore overlap and consistency here.

[timbl]

Certainly Projects are an interesting place to look at building more tools around. A github organization, the solidproject.org unincorporated organization, things which run themselves on the web in a solid pod are (very) interesting too, but I don't think of them as ResearchOrganizations .

[dr-shorthair]

A Project is usually time-bounded, more like a prov:Activity, compared to an organization which is ongoing.

I had a go at designing a project ontology based on Prov a few years ago.
The key features were that a Project has a Plan and a Budget.
See https://dr-shorthair.github.io/project-ont/docs/

[danbri]

Time-bound projects are time-bounded. But even research projects with funders often come and go unpredictably - you only know in retrospect (failed bids, staff changes, making up new names and jargon to sound more fundable, etc.). And then there are open source projects and so on. Some go on to become more institutionalized, some fizzle, some fragment, but many continue indefinitely. And many, perhaps most, have no money or budget. It sounds like you are talking about s more restricted notion, … a managed, funded. scheduled project. When these go well, do they necessarily cease to exist at the end of the initial plan? Your definition feels a little brittle on those aspects for it to be the the definition of the overall class of Projects…

…

On Sun, 18 Jul 2021 at 06:30, Simon Cox ***@***.***> wrote: A *Project* is usually explicitly time-bounded, more like a prov:Activity than an organization. I had a go at designing a project ontology based on Prov a few years ago. The key features were that a Project has a Plan and a Budget. See https://dr-shorthair.github.io/project-ont/docs/ — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#2877 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AABJSGO3S2ZIUHHIBWI5Y7LTYJRILANCNFSM44QWXKAA> .

[dr-shorthair]

It sounds like you are talking about s more restricted notion, … a managed,
funded. scheduled project.

In the ideal case yes. But OWA means that if some of these specific aspects are not known then these properties do not have to be included.

What I was trying to do was to figure out what are the typical and essential properties of 'project' in contrast with an 'organization'. A 'project' is definitionally transient, while an 'organization' is more often ongoing. The best projects have plans and budgets and leaders, and maybe all projects should aspire to these. I certainly agree that any of these can come and go during a project, but I think these characteristics are generally applicable - certainly with both formal and informal projects that I've been involved in.

Not trying to box anyone in here, just write down the usual expectations. Information/data modelling requires us to make these kind of choices sometimes, else we can't classify our world ;-)

[sharppaul]

It sounds like you are talking about s more restricted notion, … a managed,
funded. scheduled project.

In the ideal case yes. But OWA means that if some of these specific aspects are not known then these properties do not have to be included.

I feel like I'm in the same boat here. I'm working on making linked data available for the projects listed on edmerp.seadatanet.org and can't find a way to publish any temporal data for a Project.

For the SparQL ontology we used prov:Activity aswell, using prov:startedAtTime & prov:endedAtTime to publish the start and end date.

Maybe schema:Project could use something like schema:CreativeWork's temporalCoverage

Cheers,

Paul

[danbri]

Ah, looks like I didn't close the loop:

It's in schema.org now, in the Pending area awaiting feedback from implementors

https://schema.org/ResearchOrganization

@timbl - did this prove useful in the end?

[github-actions]

This issue is being nudged due to inactivity.

[fthobe]

@gmackenz can be closed