Integrate check-bib-for-predatory

[koppor]

https://github.com/CfKu/check-bib-for-predatory

See also https://discourse.jabref.org/t/double-check-the-bibliography-for-predatory-publishers-and-journals/1738

[milescajus]

Hi, just to let you know in case you weren't notified, the draft PR for this issue is now ready for review :) thanks!

[xuanan20020]

Hi, koppor, I hope you are doing well. I have noticed that milescajus is no longer interested in this issue. May I take over as an assignee. Thank you :D!

[koppor]

@xuanan20020 Sure thing. You now even have a "template" for MV file generation at /src/main/java/org/jabref/cli/JournalListMvGenerator.java#L19, which was introduced at JabRef#10298. With the previous try and that PR, you should have a good start!

[reox]

I saw this feature in 5.12 and tested it. However, I fear there are many false positives. For example, if the Publisher is "Springer", "Wiley", "IEEE", "CRC Press" or "Nature Publishing Group", the article/book gets flagged. Sure, you can have some opinions about these publishers but I hope that these are still credible :D
It is also funny that it shows "Science" or "Bone" as predatory journals - which I hope are not...

It would certainly help to show some reasoning behind the flagging, i.e., "shows up in List xyz"

[fofusa]

Hello,
My experience is the same than reox.
I just tried this feature on my own database and there are way too many false positives including Nature, Science or journal of the OSA (Optical Society of America) like Optics Letters.
These are highly respected journals that most scientists in my field would not consider predatory.
I think there is something wrong with the list used for flagging.

[koppor]

@reox @fofusa Thank you so much to replying to the original feature wish.

Can you comment on https://discourse.jabref.org/t/double-check-the-bibliography-for-predatory-publishers-and-journals/1738/3 for ideas how to refine the checker?

I think, the thing you experience is that (some) predatory journals re-use the title of existing journals. And JabRef "just" checks the title of the journal and does not do a DOI lookup to find out the publisher etc. See https://github.com/stop-predatory-journals/stop-predatory-journals.github.io/blob/master/_data/journals.csv for the full list of journals.

[Siedlerchr]

Maybe we also should limit it to full titles as we currently do also similarity check for the title ?