<institution content-type="XXXX">

[Melissa37]

eLife is upgrading to JATS and we're using the JATS tagging for FundRef information. Our xml is proposed as: <funding-source><institution-wrap><institution-id institution-id-type="FundRef">dx.doi.org/10.13039/100000002</institution-id><institution content-type="university">National Institutes of Health</institution></institution-wrap></funding-source>

According to the JATS guidelines: Common Practice: All levels of a multi-tier organization are typically listed within a single element. For example, a program, a department, and a university may be part of the same . Similarly both a corporation and a division within the corporation might be inside one . Alternatively, a divisional distinction such as an “organization” element with subordinate “dept” and an “office” inside a department could tagged using multiple <institution> elements with the @content-type attribute differentiating institution, department, and office:

<institution content-type="edu"> However, the only content-type= descriptions I can find are university, dept and edu. I am sure there will be more and also there is inconsistency there between full spelling and abbreviations.

Looking at the FundRef ids, they tag each funder with a "fundingBodyType" and "fundingBodySubType". Some of these are obvious, some are ambiguous and some could require both pieces of information to be useful.

I don't think it will be useful for eLife to use this attribute and use a bunch of our own made up terms, or just fudge in the FundRef terms if this will not be widespread.

Is this something that we of the JATS4R group are interested in addressing?

[ostephens]

This is definitely of interest to me as far as the work I'm currently engaged in for Jisc for UK Universities as the issue of who provided funding relates to monitoring whether a publication fulfils the requirements of that funder (e.g. in terms of licensing/open access)

Haven't got time this week to respond in more detail regarding the questions/issues you raise here but I'll try to do this before the call next week. For now just saying definitely of interest to me.

[Melissa37]

Great! Look forward to hearing more from you. I suspect people like Ringgold provide good classification, but this is proprietary. I wondered about ISNIs being another place where we could link in information. Likewise, CrossRef might be interested in working with us.

[Melissa37]

The short answer to this is, leave it to a future date. The longer answer is below (extract from author and affiliations sub-group meeting):

We discussed that very issue in one of our calls. Here is the excerpt from our meeting notes: "We talked about whether we need to recommend that people use @content-type on to indicate the level/type of institution, as we are aware that the JATS tag library contains examples of this. We decided that, at this time, the use of this attribute would not be helpful for reusability/interoperability because there are currently no standarised values, and since this element’s use is relatively new, it would be difficult to gain consenus at this time as to how and when to apply @content-type on this element. We also decided that when a publisher uses institution IDs, the type and level of institution should be determined by the scheme of the ID-assigning-authority. We therefore removed any instance of this attribute on from the examples. "