Describe the problem you would like to solve
Currently, when authors submit articles in OJS, they have to manually separate each keyword when entering them into the metadata form. This can lead to errors, inconsistencies, and requires additional time for data entry. Authors often copy and paste keywords from other documents, and without automatic processing, the system may retain unwanted punctuation marks or merge keywords into a single entry, complicating further search and indexing.
Describe the solution you’d like
It is proposed to add an automatic keyword processing feature in OJS. This feature would automatically split keywords entered with delimiters such as commas, semicolons, and periods. When an author enters keywords in a single field using any of these delimiters, the system should detect and process each word or phrase as a separate keyword. Additionally, the feature should remove any extra spaces around each term to ensure proper formatting. For example, if an author enters “Vocational Education, workplace training; career, profession. ICT,” the system should process and save them as individual keywords:
Vocational Education
workplace training
career
profession
ICT
This automation would simplify the submission process, reduce data entry errors, and support more accurate metadata handling.
Who is asking for this feature
This feature is primarily requested by journal administrators and editors who handle article submissions and metadata, as well as by authors submitting articles to journals using OJS. Technical support teams would also benefit, as this feature would reduce the number of metadata errors that require manual correction.
I’m generally an AI skeptic but classification of submissions is an interesting use case.
There are a couple of active pieces of work that speak to this:
#1550 Allow a journal to define a limited set of allowed keywords and reviewer interests: This would allow a journal to pick a fixed vocabulary for submission keywords
In my experience with Elsevier journals with fixed keyword vocabulary, the vocabulary is usually quite limited and consists of very generic and broad topics. Such a vocabulary may be useful for choosing appropriate editors for the submission, but I doubt that it is of much use for the indexing purposes.
This doesn’t solve the problem, though. In very old OJS versions, the keywords, to be parsed correctly, had to be separated by semicolons. This created some confusion among the authors and the editors who cannot read and thus cannot follow the hint below the input. They preferred commas. However, the problem was easily solved by another editor using search-and-replace.
Now, the authors and editors just paste the whatever-separated keyword list into the input field, creating one large “keyword phrase”. To fix this, “another editor” should go and retype every keyword manually, which is quite annoying, to be honest.
In fact, people are even creating dirty workarounds to fix this issue.
this would be a worthwhile feature. I am dealing with this issue actively on submissions where the authors just cut and paste their entire keyword block.