Discovery of resources such as Mandatory Disclosures and Reporting / STAR

[csarven]

I may have missed this entirely but is there a machine-readable way to discover resources such as the ones outlined under mandatory disclosures and reporting or STAR ?

Has this been considered, a todo, or out of scope? (If in scope, this would be immensely valuable...)

[AlexDawsonUK]

We do have a JSON API which contains a complete unabridged copy of the WSG that can be called upon. It contains all of the resources within the guidelines (though not STAR - yet), and as its in JSON format (source) it should be machine readable for tooling or queries.

If you do have any suggestions to improve or expand machine readability / discoverability though, ideas are welcome.

[AlexDawsonUK]

As there has been no further follow-up at this time, I'll close the issue (as it's been over a week so hopefully the question was successfully answered!) but feel free to drop a comment in if it requires re-opening or add another issue if you have any further questions and we will be glad to answer them.

[csarven]

Pardon me for not being clearer. If a website owner wants to publish their sustainability reports / PPP targets, how can they be made machine-discoverable?

For instance, a consumer interested in the "Achievements and Compliance" report and data ( https://example.org/achievements-and-compliance" ) of a website ( https://example.org/ ) could look for the corresponding property ( https://www.w3.org/ns/wsg#achievementsCompliance ), for example, in HTML:

<a href="https://example.org/achievements-and-compliance"
   rel="https://www.w3.org/ns/wsg#achievementsCompliance">Achievements and Compliance</a>

or in JSON-LD:

{
  "@context": "https://www.w3.org/ns/wsg.jsonld",
  "@id": "https://example.org/",
  "achievementsCompliance": {
     "@id": "https://example.org/achievements-and-compliance"
  }
}

Similarly, different properties can be used to discover other types of high-level reports - such as impact reports, standards and policies, or impact reduction - that a website makes available. I would consider this the first level of discovery.

The second level involves ensuring that all relevant data within those reports (and any relevant data) is also structured in a machine-readable format, in addition to being human-readable.

WSG already defines many criteria for machine testability, however, it does not specify how statements and accompanying data can be found or processed. Developing a corresponding vocabulary (e.g., under https://www.w3.org/ns/wsg ) to express these would further support this goal. And, there are many other existing W3C vocabularies that can be used alongside.

This would not only make sustainability reports machine-readable but also enable widescale analysis and further development, which I presume to be a desired goal of following WSG and public access to such information.

---

Related: thegreenwebfoundation/carbon-txt-site#9 (comment)

[AlexDawsonUK]

One method worth examining may be through the Schema project.

It uses structured data in a variety of formats including JSON and RDF to provide rich semantics for a variety of objects. I've made a note for us to examine the potential for how we can seek to include more sustainability related material within the project (as it already has widespread adoption and uptake within the industry).

[hanopcan]

Hi @AlexDawsonUK and @csarven

Really great to see this discussion happening. Thanks for tagging the carbon.txt project which we're working on at Green Web Foundation.

If I've understood what @csarven has raised, then I think carbon.txt could be of help in solving parts of this problem.

Certainly it can help in defining a standard place ('/carbon.txt' or '/.well-known/carbon.txt' on any URL) to share mandatory disclosures and information as a starting point. See https://carbontxt.org/ for all the detail, a short presentation here and https://thegreenwebfoundation.org/.well-known/carbon.txt for more info.

What gets shared in a file and the machine-readable part is an interesting point to explore. carbon.txt has been written to be extended for parsing any kind of structured data in theory. So far we've a working example to make carbon.txt read CSRD (Corporate Sustainability Reporting Directive) reports written in XBRL. We're hoping to get lots more plugins written to parse other kinds of structured data.

I love the idea of seeing if we can find a way to leverage carbon.txt for the benefit of the problem outlined here, and potentially explore what structured data could be exposed in this way.

If either of you are keen to discuss and bounce around some idea let me know - would love to!