I wanted to share that PKP has a new guide for better practices in journal metadata. It is available here: Better Practices in Journal Metadata.
This particular version exists in the PKP documentation hub where it can be consistently maintained and updated by the PKP Documentation Interest Group. The objective of the guide is to improve the quality of metadata registered with us by OJS users.
Hi. Iāve suggested specific documentation improvements directly to PKP (here and here).
I just wanted to share a caveat for fellow editors working with multilingual metadata in OJS.
Currently thatās only for human interaction, as itās not supported by CrossRef nor Google Scholar; plus, itās not normally exposed by OJS, so itās also not indexed in regular Google Search either. I only stress this point because it took me several years to realize multilingual metadata does not necessarily improve discoverability, unless you design a custom theme to display both languages simultaneously (see here). Fellow editors should consider if the additional effort is worth it.
The above applies to multilingual metadata of full-text content published in a single language. In case the full text is also multilingual, then publishing independent articles seem to be the closest way currently available to implement CrossRefās recommendation for depositing multiple DOIs, one for each language. The solution is not ideal, though, because the hasTranslation and isTranslationOf relation is not yet supported in OJS. But at least the secondary language content wonāt be hidden from the bots/crawlers.
Thanks for this note @fgnievinski ! I appreciate you raising it with PKP as well. Have you flagged this in the PKP Community Forum, too?
Youāre right about OJS not yet supporting the hasTranslation and isTranslationOf relations, but we do have documentation on how to achieve this via XML, if that can be an option for you/others. You can learn more about how to add the relations using XML here: Relationships between different research objects - Crossref
I wish I could contribute with more concrete bug fixes and feature improvements, unfortunately thatās not on my horizon currently. So, I try to contribute with documentation and specification, without being too intrusive with the actual development.
Thatās a good point, thanks for the reminder. Normally we use the OJS plugin for automatic DOI registration and deposit. Now I assume weād have to extract the two XML files, edit them to include the missing relations, and finally upload the modified files via web deposit form manually. Weāll give it a try, if you have any full examples (a tutorial would be great), please let us know.
If you can send me the details of the two DOIs in question (which is the original and which is the translation), I can mock up an example in this thread, assuming you are okay with me sharing this publicly?
Since youāll be sending us the XML itself, youād submit it to us using the admin tool at doi.crossref.org by following these steps:
But itās still published with two galleys in a same landing page instead of two separate landing pages.
As Iāve mentioned a while ago, the editor-in-chief is still reluctant to duplicate the bilingual articles, as heās afraid it might decrease the journal impact factor: spreading a fixed number of potential citations over a greater number of published articles.
Unfortunately, Iād need two DOIs to properly mock this up, Felipe. Since I donāt have that, Iāll hold off on the tutorial. I understand the concerns about citation splitting.
If/when you (or, anyone else reading this thread) have an example, let me know and Iām happy to go through step-by-step instructions with screenshots on this thread.
Dear @ifarley dear @fgnievinski
To a very far extend I support your suggestions. In addition, I would propose to collect and share code modifications of the original OJS CrossRef plugin. A lot of code develoment is done especially at University presses using OJS. And as to my knowledge relatively little of these customisations are shared e.g. on github or are even proposed to the core project.
So, letās start collecting. You may share your code to github .com/klausru/crossref-ojs
Or shall we start a new thread for this?
All the best
Klaus
Editor in chief and developer of www. medienpaed .com
Thanks for the link to github.com/klausru/crossref-ojs. If youāre wanting to have a discussion within the Crossref forum about suggestions or customizations, feel free to start a dedicated thread here. I fully support that idea!
On the other hand, Isaac offered a mockup of the XML files, which would be useful to guide the implementation as well; heād just need a pair of articles to demonstrate the process.