Open sylvaticus opened 4 years ago
This is already part of what our authors do, the tags include domain labels which are shown on the paper view page (e.g. https://joss.theoj.org/papers/10.21105/joss.01834) and https://joss.theoj.org/papers/tagged/exoplanets .
Then one can choose an open classification (tags) or a predefined hierarchical classification (there are many).
I've yet to find one sufficiently complete for JOSS but would love to be wrong. Could you suggest a few that you think might be sufficient?
I'd recommend to use two knowledge organization systems:
1, a broad, generic classification for browsing (such as
For the second I recommend to use Wikidata. For the first there are several classifications and all are opiniated. Just stick to one of them. Candidates:
Why not use Wikidata only? Because the articles will get fragmented and there is no clean, stable hierarchy. In terms of information retrieval: better recall.
Why not use only a broad classification? Because every classification is too broad in specific areas. In terms of information retrieval: better precision.
With the aim to improve the browsing of the many papers, I think the classifications of the papers should include a dimension relative to the application area of the paper, not just the programming language used.
Then one can choose an open classification (tags) or a predefined hierarchical classification (there are many)..