How to publish back issues (pdfs)

[geirrosset]

I am struggling with what I thought would be a no brainer.

We have 10 years worth of journals that we want to publish as Open Access. They are all divided up into separate articles like 2009_1_1.pdf and so on.
How do I get those online. I used the file manager to upload complete jopurnals and separate articles, but I can not view them anywhere except the file manager.
How do I make them "go live" som visitors can read them? I just assume that as site administrator I can publish past issues and present issues without needing to involve the editorial staff (there is no way I am going to be able to teach them to use this system, so we need to handle all publishing. The advanced editorial functionality of OJS is not of interest as we have a hard enough time getting contributors to hand in properly formatted Word files, throwing OJS into the mix would mean that we stop being publishers and start being technical support staff).

Sincerely
Geir Rosset
Novus press

[ramon]

Hello Geir,

Searching the forum for similar keywords will return quite a few examples of how to solve your problem.
The file manager is a utility that will not help you much in this case.
It seems you don't have your past issues in a database, so writing the XML file to import those issues may be quite time-consuming.
I would suggest adding the Author role to your profile, as it is the only one that executes the "submission" process.
Then, if you are registered as Journal Manager/Editor, you can use the expedite function to skip the review process.
All you'll need to add is the metadata for each article, which you may already have through cut-paste from the PDFs, or will have to type if they are not available as text.
You won't need to register previous authors to perform this task, but may have to add the <![[CDATA]]> code to the email field in order to include authors without emails, although 10 years ago, email was quite common...

In our experience, a 16-article issue is usually done in 3 to 5 days using this process.

As past issues can be added at any time, organized and dated accordingly, it is just a matter of patience until you have all your articles up.

[geirrosset]

Ok. That's a deal breaker then. Up to several days per issue, times 4 issues per volume, times 10 years of back issues, times 12 for the number of journals we publish.
So does anyone else know of a way to just get these journals out there so that they are searchable by Google Scholar? Our choice would have been to just dump them in a folder on our web server and let Google index them. Using OJS seems to me like trying to shoot sparrows with a cannon (for our purpose). A far more complicated and powerful tool than we need.

[ramon]

Hello Geir,

OJS IS by definition a "journal management" software. It allows a Journal Manager to put his journal online, allowing authors to submit their work, editors to choose the appropriate reviewers, control the review process and then publish the content in an organized Table of Contents, with additional tools for indexing, disseminating and making publicly available high quality content to a large community.

The approach of just publishing PDF's online is not actually considered "publishing" a journal, as there are several rigorous procedures and policies that need to be taken in consideration to actually publish a journal, depending on the quality your journal seeks and where it wants to be indexed.

You can, if you have the skills, manually write the XML file to import all the data required for the previous issues to be indexed by the system, allowing users to search your journal's past content, as well as index your journal in other specific journal databases, as well as Google. OJS has plenty of documentation, and this forum has many posts regarding this XML import approach, which may seem daunting at a first glance, but may save you more time. Plenty of people in this forum may be able to help you. I've used the import tool in the past, so my XML files may be outdated. I haven't tested them against the current version.

If your PDFs don't have metadata, and the data is not in a machine-readable format that can allow some sort of computer program to automate at least part of the process of importing back issues into OJS, you will eventually have to type/copy-paste the data manually, anyway. If you have to type manually, I personally suggest using the system itself as it is less error prone than writing long and "complex" (particularly structured) XML file. It will also help learning the software.

Another option would be to try and develop a PDF scraping tool, if your files have a standard format, and have that create the XML file for you.
This would take probably more time than both approaches I've mentioned.

If you are not using OJS' journal management capabilities, then OJS may not be suitable to your needs.
A CMS tool like Drupal, Joomla!, Plone, or even Wordpress may be more effective.

[geirrosset]

I have been playing with the expedited submission alterative. I have a few questions..

Why is there no easier way of adding back issues? Why do I have to create a bogus author and a bogus editor and pretend I am submitting an article to a new journal, and then pretend to edit it and pretend to proof read it, when in fact the article is ten years old and is as finished as it will ever be. There should be a much easier way for a super user to just publish old issues.

That being said, it does work.

Now for my second question.

Our back issues are all in full text searchable PDFs. Why do I need meta data? And is this meta data the same meta data that I can enter in Adobe Acrobat under File->Properties? Or is it meta data in OJS? Title and author is easy. But How would I know what the specific subject for a preface is (your average article has a title and you can always put that in there as the subject, even though that might not be entirely correct)? Contacting all the authors of all the articles from the past ten years to have them come up with keywords and subjects is an impossible task.

And finally...

If I enable Metadata Harvesting by visiting the link found under site administration, will the PDFs that I publish be made searchable through Google Scholar? (A lot of the researchers we work with use Google Scholar and that is all they are really interested in).


Addendum

There should be an OJS address book for all contributors (not necessarily available externally, but for the site admin it would be very handy). When publishing 10 years worth of journal articles there are a lot of peaople who have many articles. I now have to fill in their details over and over again. When they appear I might have forgotten where I found their details last time and so have to spend a lot of time searching for the same information...

[ramon]

Hello Geir,

Although I'm not part of the PKP Team, as a community member I try to assist as much as I can, in order to relieve them of some of the more "simpler" and bureaucratic tasks.
Having said that, your issues are all pertinent.
There is no "easy" way to publish back issues, as this cannot be fully automated, especially if you don't have the data in a database which would allow to create the XML to import into OJS.

Why is there no easier way of adding back issues? Why do I have to create a bogus author and a bogus editor and pretend I am submitting an article to a new journal, and then pretend to edit it and pretend to proof read it, when in fact the article is ten years old and is as finished as it will ever be. There should be a much easier way for a super user to just publish old issues.

Addendum

There should be an OJS address book for all contributors (not necessarily available externally, but for the site admin it would be very handy). When publishing 10 years worth of journal articles there are a lot of peaople who have many articles. I now have to fill in their details over and over again. When they appear I might have forgotten where I found their details last time and so have to spend a lot of time searching for the same information...

One way that we tried working with past issues is actually registering previous authors ourselves, as they tend to submit multiple articles to the same journal over the course of the years. This way, even though it may seem bureaucratic, you can submit all the articles of the same author once,if you have them all separated and ordered this way. There may be issues when the author is not the primary author, but it can be overcome easily. You won't be able to use the "Quick Submit" feature, but, you can skip any steps that you don't require during the process.

In the Editing stage of the submission, once approved in the Review (if not using the Quick submit/Expedite Function), is where the final document is normally submitted for publishing as a Galley file. All the other steps can be started/finished without actually performing them.

A default/standard password and saving that in your browser will save some time, but most will have to be typed/copied-and-pasted in.

Of course, you will have to come up with a way to organize your TOC's accordingly.
We created all the past issues first, which enabled a faster scheduling.

You could also disable (remove) all the checklist items during this back-issues publishing process.

Our back issues are all in full text searchable PDFs. Why do I need meta data? And is this meta data the same meta data that I can enter in Adobe Acrobat under File->Properties? Or is it meta data in OJS? Title and author is easy. But How would I know what the specific subject for a preface is (your average article has a title and you can always put that in there as the subject, even though that might not be entirely correct)? Contacting all the authors of all the articles from the past ten years to have them come up with keywords and subjects is an impossible task.

Metadata are required, in order for OxS to provide a precise search result, not only full-text search. The Metadata provides a form of classification that full-text does not provide, and the precision comes from specific fields, not just keywords, according to the Dublin Core OAI format. Also, the mandatory Metadata fields in OAI-DC are just Title and Abstract. All the others are optional and can be updated at any time or turned off.
You don't even need to submit a document during the submission process itself, only in the Editing stage.

And finally...

If I enable Metadata Harvesting by visiting the link found under site administration, will the PDFs that I publish be made searchable through Google Scholar? (A lot of the researchers we work with use Google Scholar and that is all they are really interested in).

I cannot comment much on how Google Scholar works. Google itself will harvest anything you have, specially if you are in a domain that's already indexed by it.
Otherwise, you will have to wait for Google to work its magic.
The Metadata Harvesting has nothing to do with Google Scholar, as far as I know, as it is not (I believe, can't make the affirmation) the way Google indexes things. Metadata Harvesting in the context of OxS is in reference to the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH).
Also, in order for OJS to index the full-text of PDFs, you will need to install a server-side application (the forum has quite a few posts on this subject such as this one), as described somewhere in the documentation.

[asmecher]

Hi geirrosset,

Two notes to add: it doesn't sound like you've tried the "Quick Submit" plugin, which should be available to the Journal Manager via "Import/Export Data". Also, regarding Google Scholar, any recent version of OJS will automatically expose metadata to Google Scholar as long as you haven't put the article abstracts behind a login.

Regards,
Alec Smecher
Public Knowledge Project Team

[geirrosset]

asmecher: I have actually used the QuickSubmit plugin. I found it yesterday by chance. But thanks What you say about Google Scholar sounds good.

I have another question about meta data. Where do I input the metadata for a particular article? I haven't seen any metadata or keyword fields other that title and summary?

And finally
What about books? We have books that are sold out that it would never pay to have re printed, but that people still ask for every now and then. Can we publish books using OJS or is there a similar system for books?

Geir

[asmecher]

Hi Geir,

Check out Journal Setup, page 3, to enable or disable the various metadata fields.

OJS would work fine for books, off the top of my head; it's somewhat clumsy to use the OJS workflow for putting together book content, because treating each book as a single word processor document is often too cumbersome, but in your case the content is already complete and you're just looking for a place to host them. I think that should work OK.

Regards,
Alec Smecher
Public Knowledge Project Team