What is the accepted percentage of overlap in the expanded conference version of a journal? [duplicate]

Question

I submitted a journal paper to a Springer journal (Computer Science), and it is still under review for a year. I submitted the fourth revision recently. As it is taking more time than expected and I wanted to get some publication soon, I extracted a part of it and submitted to an IEEE conference. The conference paper got accepted now and published at IEEExplore. Now when I check the overlap between my journal submission and the conference paper using some online tools I can see that they have 14% overlap (including all the references). But my entire conference paper is from my journal without rephrasing. So if I calculate the overlap in the other way I can get over 70% of overlap.

So my question is what will happen to my journal submission? Even if it is accepted will it get rejected at some point (may be at the publishing stage)? What can I do to publish it without any problem?

Answer 1

Note. I am not a lawyer, so the advice is conservative. But self-plagiarism isn't normally a legal issue. Copyright is, however.

If you don't cite either in the other then you can certainly be accused of self-plagiarism. I doubt that either Springer or IEEE would be happy with you if they notice. You have the additional problem that both will probably want copyright and you will have compromised that unless you have an explicit license from IEEE for this.

The proper way to handle it is to treat one as primary (probably the journal) and the other as a derived work. In the derived work you quote only sparingly and cite as necessary. And, paraphrasing (without citation) doesn't really get you home free.

The purpose of avoiding self plagiarism is that the reader of a scholarly article may want the full context of what they are reading including all prior works along with the context (references, say) from those prior works. What you have done seems to break that chain of context. This is part of the more general concept of plagiarism but still important.

I suggest that you find a way to deal with this explicitly. You could, logically, I think, consider either version to be "primary" and then adjust the other to treat that primary version in the same way you would treat a paper by a different author: quote sparingly and cite. And cite even for paraphrased ideas. You should probably contact the editor of the journal.

This would be a bit easier if you were dealing with a single publisher, but the problem doesn't disappear even then.

But, without citation, the "accepted overlap" as you describe it is just about zero.