Modelling adjacent data: funding opportunities, grant programme modelling

[mrshll1001]

From https://trello.com/c/gMEHjz31/33-adjacent-data-funding-opportunities-grant-programme-modelling

This came up at an ODSC team meeting. Is now the time for us to start looking at modelling other concepts that are adjacent to grants?

Options include:

• funding opportunities / grant programmes

• info about funders, such as financial year periods

[mariongalley]

Is there a user need for this? I imagine if we did some user research with fundraisers they would likely say they want to know more about grant programmes, although currently the main issues there are lack of completeness and duplicates. But there are other links to things like pooled/match funds and other types of regranting.

[drkane]

A (long) while back I had a go at what a possible extension along these lines could look like: https://gist.github.com/drkane/f47993dc75e2a8047ae9928c9673e477

At the time the user need was how do we get a view of all the available funds that doesn't involve looking at all the websites of and/or phoning grant funders to find out the information. I think there are technical barriers to my proposal - having a CSV with all your grants is relatively easy compared to keeping up to date a json file with all the details of your available funding in it.

Potentially this is an area where LLMs have made it easier to parse the information available on funder's websites - I don't know of anyone doing this but I imagine it exists. I think it could lead to issues around automating the application process - automated opportunities research + automated application writing and submission = a large number of low-quality applications.

I'm not sure in this case whether a data standard would help or hinder this issue, although I think something that codifies the "hard" criteria for a grant funding programme (e.g. type of organisation, latest income, amount applied for) could help stop applications that are ineligible.

[mrshll1001]

Is there a user need for this? I imagine if we did some user research with fundraisers they would likely say they want to know more about grant programmes, although currently the main issues there are lack of completeness and duplicates. But there are other links to things like pooled/match funds and other types of regranting.

I'm not sure on whether there is a strong user need for this.

FWIW, I would be against modelling more adjacent data in the Standard. I think it's important to draw a line around what's reasonable for our scope. I already think we should remove all the Transactions if we get the opportunity for a MAJOR upgrade.

For Funding Opportunities, CommonGrants have spent some time modelling this so we'd be duplicating effort if we attempted this.

• CommonGrants: Opportunity model

The publication patterns for Funding Opportunites would likely be totally different and I wouldn't want to end up supporting infrastructure which became a de facto national database for searching for grant opportunities.

This isn't to say I don't want standardised data around Funding Opportunities; but I don't think 360Giving is the standard to do it right now and we should be focusing on servicing our own use-cases as best as we can.

If there is strong enough demand in the future for us to support Funding Opportunities, I'd seriously recommend us looking CommonGrants as a first port of call there. We'd need to determine whether the model was sufficient for the use-cases we need, and then see if the model could support the types of publication pattern and tooling our userbase relies on.

For Grant Programmes; I think this is within scope of 360Giving (it makes sense to me to declare whether a grant was made as part of a Programme) but we can determine how much detail we want to "in-house". I'd be very curious to know what more people want to know about Grant Programmes.

The current model of a GrantProgramme in 360Giving has: code, title, description, url, and dateModified properties.

I think this is sufficient for us now, but I'm very open to expanding this based on user feedback. In an ideal world, the url field would be all that's required in a linked data approach as a Unique, Persistent, and Resolvable identifier for a Grant Programme, which could then be followed for more machine-readable information on Grant Programmes.

In the real world, I'm happy to think about what people need and whether that's reasonably in scope for the Standard or whether we want to keep with our current practice of a headline description of which programme the Grant was given as part of.

For "info about funders, such as financial year periods", I'd also be against this based on scope. Unless there's a really strong case for Publishers to declare this in their 360Giving publication, I think this is best handled by existing services such as the Charity Commission. With a good enough identifier in the publication, the 360Giving datastore could retrieve this and enrich the original grant data as it passes into our tooling.

[drkane]

I hadn't seen CommonGrants (or if I had I'd forgotten it) - I think that does a good job of this issue.