6

My papers were rejected from a top conference venue in computer science due to laughably fallacious arguments. The reject decisions were based upon the following comments:

  1. The one's complement checksum of an all zero bit sequence is zero!
  2. O(n2) can be more efficient than O(n), because for n = 0.9, 0.92 < 0.9 !

What should I do now?

17
  • 13
    Note that, although comment 2 is expressed in a ludicrous way (we only care about positive integer n, and the smiley for crying out loud), there is a significant point there. A linear time algorithm is not necessarily faster than a quadratic time algorithm for inputs that we care about. For example, 10n^2 < 500n for all n < 50. – David Richerby Jul 23 '16 at 11:02
  • 10
    "O(0.9) is a meaningless notion" That's not correct. It's the same as O(1) [constant time algorithms]. – FooBar Jul 23 '16 at 11:40
  • 29
    @Christof You seem extremely defensive about anything that you perceive to be a criticism of your work. I never said anything at all about the quality of your contribution. I couldn't possibly say anything about the quality of your contribution because I have no idea what your paper says. All I said is that, in practice, a linear-time algorithm is not necessarily faster than a quadratic one. That is indisputably true. It is a point that one of your reviewers made. You should consider whether it is a relevant point, even though your reviewer said it in a ridiculous way. – David Richerby Jul 23 '16 at 11:47
  • 8
    @Christof Furthermore, the call for papers says "Theoretical papers must make a convincing argument for the relevance of the results to secure systems." So, actually, reducing the complexity on its own doesn't seem to be a fair contribution to this particular conference unless you can argue that it's practically, as well as theoretically relevant. – David Richerby Jul 23 '16 at 11:53
  • 10
    Even if you believe that the reviewers are idiots, there is no appeal phase in most CS conferences once the decision is final. BTW this conference had a rebuttal phase, which already gave you a chance to state your claims. It did not work before. Just let it go. – Alexandros Jul 23 '16 at 13:27
23

What you need to do is to understand why your paper was rejected. I'm pretty sure your paper was not rejected from what you describe as "a top CS conference" for such obviously false reasons. Reviewers don't decide whether the paper is accepted or rejected: they advise the programme committee. The programme committee evaluates the reviews before coming to a decision and a review containing such obviously nonsensical comments will carry very little weight. If your paper was rejected because of the reviews, it was the other reviews, not this one.

Note also that, at top conferences, many papers are rejected simply due to bad luck. Usually, there are many papers that are a good enough quality to be accepted but the conference isn't big enough to take all of them. The actual reason for rejection may just be "We liked your paper and it's of essentially the same quality as papers we accepted, but there just wasn't room for yours and we liked these other ones just slightly more."

7
  • My paper was rejected from ACM CCS. And the review comment was only that one single line (plus some biased quantifiers), even after rebuttal. – Selestine Jul 23 '16 at 10:31
  • 2
    @Christof The nonsense review was not the reason for rejection. You received other reviews, right? What did they say? – David Richerby Jul 23 '16 at 10:42
  • 5
    You believe your paper was rejected for reason X. It was not rejected for reason X. I suggest you discuss it with your advisor or with a colleague who's familiar with your paper. – David Richerby Jul 23 '16 at 11:00
  • 36
    The best act is to understand why your paper was rejected. – David Richerby Jul 23 '16 at 11:14
  • 2
    As a conference reviewer, I have seen co-reviewers do sloppy work. Very sloppy work.It's the job of the PC to keep quality control on that. If they do not do it, and let clearly inane reviews through on a systematic (occasional slips happen!) basis, it's not a good conference, or it's on the way of becoming one that's not good. Look for a better one. – Captain Emacs Jul 23 '16 at 18:08
18

It sounds like you've already gone through the rebuttal process, so as far as I know, there's no further appeal. I'm going to focus on what seems like the core issue: what do you do if your paper is rejected for reasons that are utterly stupid?

  1. Take a deep breath (or whatever your preferred calming action is). Rejection is a big part of being an academic; it's never pleasant, but learning how to handle it well is absolutely necessary to survive. That doesn't have to mean shrug it off; have a coping mechanism for handling rejection, and use it. (As an aside, I think it's something that doesn't get discussed enough when talking about career choices; I've known people who would have been fine researchers and teachers, but would have been destroyed by how much they had to deal with rejection.)
  2. Take some time (a day or two, or a week) away from the issue to get some distance. During that time, feel free to be furious and vent (to appropriate audiences) about how stupid the reviewers and PC were.
  3. When you're ready, come back to the reviews. Retread them carefully; it's quite easy when you're angry to misread a comment that gives important context as a vapid criticism. This time, read as charitably as possible. Consider that you might misunderstand something important; consider that they might have said something unclearly; consider that their comment is confused because they're confused about something in your paper. Someone in your audience read your paper and had this response; they may well have misread your paper in a stupid way, but even so, that's something you need to fix, even if it seems blindingly obvious that they're wrong, because if one person had that misunderstanding, it's quite likely other readers too.
  4. If you really can't make sense of the review, ask a colleague. Don't ask them "can you believe how dumb this review is?", ask, "I'm really not sure what this review is saying, could you give an outside perspective?" If your colleague can't make sense of it either, then you can throw up your hands about it. The reviewer really did screw up, either by having stupid comments, or by writing them so unclearly that they can't be deciphered.
  5. Revise, resubmit, and move on. Rejection is part of the lifestyle.
2
  • 4
    +1 for the second half of point 3 - this is vitally important and far too few people see it this way. Same thing for people writing furious rebuttals as letters to the editor explaining how the referee got it wrong: the clarifications need to go in the text of the paper, where they can prevent others in the audience from making the same mis-reading as the referee. – E.P. Jul 23 '16 at 20:31
  • 2
    As a way to help the OP get some perspective on the problem, I recommend shitmyreviewerssay.tumblr.com. After a few good laughs, you may be able to read the reviews of your paper with less emotion. – mhwombat Jul 25 '16 at 17:03
5

First, Wrong decisions are made everywhere, everytime in the academia. Calm down.

Second, as far as I know about CS conferences, once a decision is made, there's no chance to appeal against a rejection. However, you may want to send your paper to another conference. Eventually it'll be accepted. Sometime, somewhere.

2
  • That is correct, but I feel I can't choke on it. the review comments are not only unconstructive but are also fallacious. This makes me a lot of doubts in the review system and the role of pc members. – Selestine Jul 23 '16 at 10:28
  • 15
    Roll your eyes. Shake your head. Invent a not-so-nice designation of those reviewers. Then, move on, and find someone else who can appreciate your work more. – Guntram Blohm Jul 23 '16 at 13:19

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.