12

My question is simple: If you have a paper that got rejected two or three times (border-line* rejects every time) from top-tier conferences, would you polish it and resubmit it again to similar or another high rank conference or would you try some lower rank venues ?

* Border-line reject: every time got 2 good reviews and 1 bad review.

P.S. Every time the paper got vague suggestions like: compare with this or that. And the first and second time we DID in fact address all concerns, but another concern showed up next time.

P.P.S. I am concerned with CS conferences.

Thank you.

Improve this question
1
  • It depends on the reviewers and the research problem. If they note that the problem or findings are not significant, then go to another venue, assuming they are correct. If not, then revise your arguments. However, if there is an issue with your execution/validation/presentation, then I would continue to re-submit the paper after revision.
    – Prof. Santa Claus
    Mar 22 at 5:46
24

I would not continue submitting the same paper to top-tier conferences after three rejections. The Universe Has Spoken. Move On.

Many of my colleagues do submit such papers to lower-tier conferences, but I usually just send my papers directly to journals after two conference rejections.

Improve this answer
2
  • I think my plan is to go for a second-tier conference and later try to extend the paper for a journal submission.
    – NeoN
    Dec 27 '13 at 23:45
  • I usually send my papers that I do not consider top directly to journals. Otherwise I waste everybody's time. But many times we tend to highly overestimate our work and think it deserves to be accepted to STOC etc. If a paper has been rejected more than 2 times that's a good signal to either go for journal or lower tier conference (I do not see direct purpose for that latter tho)
    – PsySp
    Dec 21 '17 at 15:36
11

There is only so long any paper, whether it be for a conference or for a journal, should be held up in the review process. While it is stuck in review, other people may be able to publish similar (or stronger) results, at which your work loses the "novelty" factor, which will make it less competitive in the future.

At a certain point, you have to cut your losses and move on. Where that point of diminishing returns for presenting in a top-tier conference is depends upon your risk tolerance. But I'd much rather get my paper presented or published somewhere within a year in a second-tier conference than wait for two years to present at a top conference.

Improve this answer
2
  • I guess I should do the same and perhaps submit to a second-tier conference. Unfortunately the conference publication game is becoming more like a random game of luck these days...
    – NeoN
    Dec 27 '13 at 23:31
  • 5
    Not only luck but politics. I look these days at lots of researcher's profiles and participate in number of talks about publications. What I learnt is that publication has 3 legs; quality, chance, and politics. However, sometimes politics is more effective than two others, especially when you see that a publisher who has couple of papers published in top-tier X conference (in the same year), was accidentally organizing member of that event!!! I don't extend such attitude to all CS researchers, but that minimal politics happens and cause you and other unconnected researchers to miss the slot.
    – Espanta
    Dec 28 '13 at 7:05
0

Do you make any fundamental improvements after being rejected? If no, resubmitting it too many times can only reduce your chance of being accepted.

This happens to me as a reviewer. I reviewed a paper in 18 and rejected it. Then I was invited to review the same paper in 19 at a different conference. The authors changed the title but not the content. I compared it against the old manuscript in my mailbox. When I noticed the authors did not make anything new out of it, I simply gave it another rejection.

The group of people working on your topic can be surprisingly small and you are very likely to encounter the same reviewer when you submit to top-tier conferences again and again. I believe they will do the same thing as I did.

But if you follow the reviewers' suggestions to improve it and/or add new content to the paper after being rejected it will be a different story. I would think of it as a new paper, and will definitely resubmit it to top-tiers if I think the paper deserves it.

Improve this answer
-2

It is a very common practice to start from submitting the paper to the high end journals but then submit to lower level journals if rejected for not being "flashy" enough. If the paper is further rejected, it may be converted into poster and presented in some conference. If it does not make into top tier conference, then still might be good enough for a small local conference.

A good professor can tell at glance where in this scale the work stands and is it possible to improve it and get higher. However professors mostly gain this experience from the known accepted and rejected submissions.

Improve this answer
4
  • This question is about computer science conferences which (unlike in other fields) are both more visible and more selective than CS journals.
    – JeffE
    Dec 21 '17 at 15:04
  • @JeffE Will in general I agree (esp. about visibility), do you think that TOP conferences in CS are more selective than TOP journals in CS? For whatever reason, I like to publish most of my work directly to journals.
    – PsySp
    Dec 21 '17 at 15:39
  • I have spend lots of time in academic environment, just this is the way things are. Also, if the paper is really bad, I expect all reviewers to be in consensus.
    – h22
    Dec 21 '17 at 16:11
  • @PsySp Yes, absolutely. Top CS conferences are more selective than top CS journals.
    – JeffE
    Dec 22 '17 at 3:31

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.