Anonymizing submissions to Janeway

By MPE

Lots of the journals that run on Janeway, unsurprisingly, use a double-blind peer review process, in which it must not be clear, to various parties, who authored various documents.

People are rubbish, though, at anonymizing their documents. Sometimes they write their details in the file (their name, for instance), other times they cite their own work, sometimes they cite grant numbers that can be traced back to them. It’s all very difficult.

Hence, it was of interest, recently, when SAGE announced a $5,000 grant to “build an open source tool that will allow authors to check if a manuscript has been properly anonymized before double-anonymous peer review”. The problem is that this was massively over-specced, in my view, for the money. For instance, they wanted to run checks that “author names, emails, author biographies and affiliations do not appear within the manuscript or supplementary information”. They wanted to “remove references to funding sources or anonymise the funding source(s), e.g.: The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was funded by [details omitted for double-anonymised peer review]”. This type of full-text search work is incredibly difficult.

We decided, in Janeway, to pursue a more limited anonymization routine that will strip identifying metadata from files and have wanted to do this since September. This works on .docx, .odt, and .pdf files (as well as images). It uses the excellent MAT2 to do its work. We can’t, sadly, handle old-school .doc files at the moment. Nonetheless, the basic version of this is now live and kicking!

When these fixes are merged into the main release of Janeway, the File History page now gets a new set of information and options that show if there is identifiable metadata in a file. You can also view a full output of detected metadata by clicking the “show full metadata dump” link. (With apologies to my colleague, Caroline, for using her teaching module handbook as a sample file!)

An image of the document management file history

The “Erase metadata” button here does what you would expect. It attempts to erase metadata in the file, and adds it as the latest revision of that file to the system. As you can see in the below image, the metadata is no longer present and the underlying file has been scrubbed clean, which is much safer for reviewers to see.

The same file with the metadata erased

I’ve then also hooked this directly into the submission system, so that new manuscripts go through automatically. From the author experience side this should be seamless: the manuscript is simply replaced by the anonymized version at upload, with the original file in the history in case of corruption. The below image shows how this appears to an author and an editor, with the cleaned manuscript presented.

A file directly after submission. Note the “cleaned” in its filename.

Overall, this was a fun Saturday hack project. We could do much more. I’d like to upstream some fixes into MAT2 to anonymize comments in LibreOffice. There are also some residual issues with extracting LibreOffice metadata that are odd… I believe it may be a dependency issue in my Docker install, but I’ll investigate that when I can. Nonetheless, all file anonymization is working.

— Martin Paul Eve

We are Andy, Mauro and Martin and we develop the Janeway publishing platform.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store