4

I know that it is a common practice in computer science to submit a short version of article in conference and extended version in journal.

What if I submit same article one short version to conference and full version to journal simultaneously?

The journal has already mentioned that they accept the extended PUBLISHED conference articles; but in my case the conference article is not published at the time of journal submission. Moreover, the review process will complete almost within same time frame for both the conference and the journal.

I also want to know what does it mean when journal say that they accept the conference paper if it is 30% extended? It means 70% of text and results can be duplicated in both conference and journal article?

  • 3
    Is the short version essentially a subset of the extended version? If so, that's double submission IMO and a no-no. – tonysdg Jan 26 '17 at 1:26
  • Yes it is a subset of extended version. But it is a common practice in CS to submit extended version of conference paper in journal. – MBK Jan 26 '17 at 3:56
  • 2
    It depends on the discipline, and under no circumstances I would say it is a no-no as @tonysdg commented. For example in Electrical Engineering it is very common to present concepts or brief ideas in a conference paper followed by an extended submission to a journal. There are even cases where, once a paper is accepted to a journal, you are invited to submit a short version to present at a conference! – o4tlulz Jan 26 '17 at 5:04
  • 2
    Actually, most CS conferences forbid simultaneous submission to other conferences or journals, even though submission to a journal after the conference paper is accepted is allowed. There's a fair chance that the set of reviewers for both versions would overlap, which risks automatic rejection from both venues. Read the call for papers carefully. – JeffE Mar 27 '17 at 12:41
  • 2
    Computer scientists lie to themselves that this is "okay." -- Publication ethics are social contracts. Computer science, or at least some subfields of computer science, agrees as a community that publishing essentially the same paper in a thesis, and then in a technical report, and then on the arXiv, and then at a conference, and then in a journal is okay. That makes it okay, despite your opinion that it shouldn't be. – JeffE Mar 27 '17 at 13:08
2

Your milage will vary on this, but I would advise against it. The spirit of the practice of extending conference submissions to journals is that one can use the conference to discuss a first version of the research with colleagues, go back to the lab, improve the work based on the feedback, and then submit a vetted version to a journal. In your scheme, if you submit both manuscripts at the same time, and they get published about the same time, what's the point of your conference paper? You won't be able to take any feedback into account, and the community also does not really gain by there being a published version of the research that is basically superseded from day 1.

I understand, and sympathize, that for you personally the advantage may well be two additional lines in your CV rather than one, but you should be aware that the academic community gets increasingly more stingy about the topic of "salami-slicing" papers, that is, the practice of trying to squeeze out as many papers as possible from the same research - and your plan, even if not strictly unethical, can easily be seen as salami-slicing, or at least be perceived similar in spirit.

I also want to know what does it mean when journal say that they accept the conference paper if it is 30% extended? It means 70% of text and results can be duplicated in both conference and journal article?

It means that there should be about 1/3 more intellectual contribution in the journal version than in the conference paper. Clearly, this is only a rough guideline, as there is no metric for "contribution". However, you should not understand this as 70% of the text can be verbatim the same. Some amount of text duplication is often tolerated for journal extensions, but 70% seems way, way above what I would tolerate as a reviewer or editor.

  • In general, I do not quite agree with the first part of your answer. I agree that the spirit should be like that, but in practice we see that conference and journal versions are virtually the same. I asked a related question recently: academia.stackexchange.com/questions/86984/… – PsySp Mar 27 '17 at 10:47
2

Submitting a paper to a CS conference and then to a CS journal (usually with extensions, though that's indeed complicated) is clearly accepted. Simultaneous submissions is clearly forbidden and punished seriously. Let me quote from the ACM policy as one example (https://www.acm.org/publications/policies/simultaneous-submissions):

The ACM does not normally permit manuscripts under review in its journals or conference proceedings to be simultaneously under review for another publication. [...] Repeated violations may lead to a ban on future submissions at the discretion of the EiC or PC.

That page does suggest this is misconduct but still threatens a submission ban.

http://www.sigplan.org/Resources/Policies/Republication/

IEEE guidelines are similar (as also mentioned in https://academia.stackexchange.com/a/18440/8966):

As described in Section 8.2.1.B.9, authors should only submit original work that has neither appeared elsewhere for publication, nor which is under review for another refereed publication. [...] Authors in violation of these multiple-submission guidelines are subject to corrective actions.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.