Does anyone here who uses Digital Commons for their journals have experience using the Crossref DOI Registration XML Generator? (Sorry, apparently I can’t add the direct link in this post? It’s from Scholar Tech dot Org) I did a search here and couldn’t find anything but perhaps I overlooked it.
My supervisor found this and I did a bit of testing it out on journals we’ve done previously, to see if it does indeed work. I’m just curious if anyone else has used this and if so, what their thoughts are on the helpfulness (or lack of) and ease of use of this tool.
I took a look at the Crossref DOI Registration XML Generator here, and tested it out a bit. Unless I’m missing something specific to Digital Commons, it doesn’t quite work.
There are a few problems:
It only produces metadata for the journal and issue, but not for any articles. Some minimal journal- and issue-level metadata is required to register articles, but there’s not much value in submitting journal and issue metadata on its own.
It’s missing a required <publication_date> element within <journal_issue>
It doesn’t include any DOIs, even journal- and issue-level DOIs. So, there’s no way to use it to actually register an identifier
All that said, if you want to enter metadata manually and produce valid Crossref XML files, we already have a tool for that. Two of them, actually.
The older one, Web Deposit Form, literally does just what this scholartech.org tool does, but more thoroughly and with more options. You enter the metadata, the form transforms it into XML (you’ll be emailed a copy of the xml file) and that XML is then submitted to our system for processing registration.
We also have a newer, more user-friendly manual metadata entry tool called Metadata Manager. It has a slightly different workflow (you start from the article and then ‘add it’ to a journal and optionally to an issue), and the nuts and bolts of the xml is hidden, but the same functionality in the end.
Please me know if you have any additional questions.
I’m not sure if you have access to Digital Commons, but this particular tool is dependent on an Excel spreadsheet that’s generated by the Digital Commons system (instruction #2). You then convert that Excel spreadsheet into a CSV (instruction #3), and then upload that CSV to the generator (instruction #4). From there, you fill out the journal info in the boxes on the page (instruction #5), hit the submit button to generate the XML file (instruction #6), open the XML file to check that the data transferred properly (instruction #7), and also check the XML file (to make sure there’s no errors) by using the Crossref Metadata Quality Check tool (instruction #9).
I’ve written to the author of this generator tool to see if they can provide the links that I think are supposed to be there (one for the spreadsheet template, one for the demo video) but haven’t heard back yet.
I do know about the Web Deposit Form and the Metadata Manager - I’ve used both, and I’m very grateful for & prefer the Metadata Manager when I’m doing just a few at a time, or even a smaller issue. Currently I’m working through back issues of one of our journals which has MANY issues, each with MANY articles and MANY authors, and going through one-by-one using either of those tools is very labor intensive, and I am not experienced enough with XML to create, from scratch, the XML file that gets uploaded using the Crossref Submission tool.
My supervisor & I were just hoping to save me some time and effort, which is why we checked out the generator.
I don’t have access to a Digital Commons instance, but if you have an example of a csv like the one that tool expects, and can send it to us at support@crossref.org, I can use that to mock up a version with sample data and upload it to test out the xml that tool produces.
Of course, I can see why it would be faster and easier to start with most of the metadata in csv format, rather than having to enter it by hand. We get a lot of requests for a csv-to-xml conversion, but it’s really impossible to make a generic one, because our schema is so deeply-nested and not-tabular, and it allows for a wide range of optional metadata. However, if the metadata coming from Digital Commons is very consistent from one article to the next and one journal to the next, it seems completely feasible to convert from csv with little-to-no data loss.