Repository navigation
Make it easy to upload anonymized data for author-blind peer review #2788
Description
Activity
@lizzieinvancouver this is the correct repo, thanks for making an issue for this!
Related to #2696, where authenticated users require access to data for review purposes (iirc).
First step for this feature is to think through the requirements:
- If we add an option to obscure the author + email on the dataset landing page (e.g. for peer review), the "this dataset should be anonymous" setting needs to be persisted as dataset-level configuration. Where does that live? An annotation?
- Even with UI anonymization, the submitter will still be identifiable in the sysMeta (
submitter/rightsHolder). Is there a way around this? Otherwise, we should be explicit about what "anonymous" means. - How do we display this option in the editor? Maybe a toggle like "Anonymize submitter and authors" (or similar) with helper text that explains example use cases (peer review) and the sysMeta limitation.
@lizzieinvancouver thanks so much for the feature request!
@robyngit those are all good points you made. I don't really see a way to allow submission without the KNB knowing who the submitter is, but I do think it would be possible to keep the submitted dataset private (to only that submitter) and introduce a new
anonymous accessAPI that could strip identifying details from the record and display that at the user request. For example, the steps might be:- Lizzie submits a private dataset
- Lizzie adds an access rule to allow
anonymized public read(this would be a whole thing on the backend) - The Metacat system generates an anonymous API access URI (e.g.,
https://knb.ecoinformatics.org/knb/d1/mn/v2/anon/{PID}) that provides an anonymized view of science metadata and system metadata, which metacatUI could display. The anonymous API would not record the identity (IP address or subject) of people accessing the record, nor expose the information of the submitter. We could only anonymize key fields in the metadata (e.g., the list of creators, contacts and parties, and maybe the spatial location, and probably only for key metadata types that we understand well (EML, ISO). The submitter would have to take care to not include identifying information in their abstract, methods, data files, etc.
Would something like that work? If so, it would be a great feature. Not sure where we'd get resources to implement this, as it is fairly involved, but I am guessing it would be a helpful feature to many of our repositories.
@mbjones and @robyngit for considering it! I don't think it reduces the work much but I do want to clarify that I doubt most of the journals asking for anonymized data will ever care if someone could track down the submitter/viewer/reviewer through the API -- they just want it to superficially look anonymous.
Brought up again by someone on the KNB, but asking for the more simple reviewer-blind review, where the author doesn't know who the reviewers are (vs. the original request for author-blind/double-blind peer-review)
- linked a pull request that will close this issueFeature 2126 layer loading indicator #2888
on Sep 28, 2026
Describe the feature you'd like
Some way to upload data without name/email identifying information of the uploader so it meets journal requirements for blind peer review. And then later add that info ... I tried doing this a while back, but could not see how to post without giving my info in some way. I tried setting my name to anonymous, but then it required an email address ...
Is your feature request related to a problem? Please describe.
This is not a problem, this is just something I wish existed.
Additional context
I pestered Matt Jones about this in summer 2024 and he told me I could submit a feature request if I wanted to -- I decided this is the repo for that (apologies if I am wrong) and finally am getting around to it now. I will be at NCEAS in person at the end of the month if anyone wants to discuss in person.