8

This question is similar to Overcoming Fear of Rejection, but not the same.

My experience in academia was teaching Computer Science for four years at a mostly liberal arts college, and that was 30 years ago. I frankly found the world of work to provide more challenging intellectual stimulation, at least in the sort of practical issues I cared most about.

There are highly technical issues in software development that I and other practitioners have discovered that are quite orthogonal (to say the least) to the trend of publishing in those areas. So when I try to publish it is hard to build much of a bibliography. When I submit (through an online submittal process) I tend to get responses that vary in quality, but are generally negative.

I am tempted to revise, in such a way as to address the concerns raised by the better quality reviews. I could buttress the claims with stronger mathematical arguments, but it's hard to tell if this would make the paper more or less approachable.

Then there's the question whether I should resubmit to the same journal. There don't really seem to be any others at the same level of relevance to practitioners.

Sorry if this is a noobie question.

Community
  • 1
Mike Dunlavey
  • 183
  • 6
  • 1
    I think it's an interesting question, but I'm not sure to see what you mean by "orthogonal issues": do you mean you found an unorthodox way to solve a known problem, or did you find a problem that nobody else is considering as a problem? –  Sep 20 '12 at 14:01
  • 2
    Being in CS myself, I find it difficult to give general advice without knowing the details. Maybe you could provide a little more detail for the CS-ers amongst us. One thing is that Software Development is not the same things as CS. – Dave Clarke Sep 20 '12 at 14:28
  • @Dave: I really wanted to keep it general, but here are some links: A, B, C, and linked posts. It's something that other practitioners have also found, but I seem to be the only one who's tried to publish it. – Mike Dunlavey Sep 20 '12 at 14:48
  • @Dave: You're right, it's only CS in the sense that CS departments are teaching something sort of automatically that I wish they would take a new look at, because their students are flooding the workplace with certain ideas I have a problem with. – Mike Dunlavey Sep 20 '12 at 15:40
  • @Charles: You asked "unorthodox way to solve a known problem, or did you find a problem that nobody else is considering as a problem". Well it's kind of both, I guess. The literature treats it as a problem having a pretty well-defined framework, for which all they can do is more deeply elaborate it, when in fact the practical form of the problem is qualitatively much different, needing a qualitatively different solution. – Mike Dunlavey Sep 20 '12 at 15:51
  • If you can show that the practical form of the problem is different than the standard form, and you can provide a better solution for this practical form, then it's not orthogonal, it's a clear improvement of the state-of-art. Nothing prevents you from having a detailed biblio on the problem, and then explain why in practice it's different, as long as you can demonstrate it. What are the main criticisms of the good quality reviews? –  Sep 20 '12 at 17:10
  • @Charles: The main criticisms (as I read them) is that I failed to make the case, and that being pointed out in detail tells me how to focus revisions. I did include a detailed use case, but it seemed disconnected from the argument. The other part of my question was if it's possible to re-submit, though they did not encourage me to do that. Your comment is helpful, and could be an answer. – Mike Dunlavey Sep 20 '12 at 17:20
  • @Charles: Another point (bear with me) is that it was suggested I do a survey of practical cases of performance tuning. I can't really do a survey. All I've got is generalities based on decades of experience and one concrete fleshed-out example posted on-line, plus rational and statistical argument. (It is a "Practice and Experience" journal.) So I'm a bit nonplussed about that kind of request. – Mike Dunlavey Sep 20 '12 at 18:02

1 Answers1

7

I certainly can't speak for all reviewers, and what I'm saying might be true only for me, but when I review a paper, I try to ask myself three questions: is it interesting for the venue? is it correct? is it improving the state-of-the-art? I need to be convinced that all answers are "yes", and the burden of convincing me is on the authors.

A problem I've seen with some papers is that they focus mostly on showing the correctness of their approach, and at first glance, it seems to be the case with the paper you're describing. Showing the interest of the approach is tricky, because often the authors are convinced it's interesting (otherwise, they probably wouldn't work on it), and they to think that everybody else understand why it's interesting. Similarly, explaining why it's improving the state-of-the-art might be tricky, especially for novel approaches, but it's necessary to explain why other approaches won't work. You have the right to write: "To the best of our knowledge, this approach has never been used to tackle this particular issue", but then, as a reviewer, I need to be convinced that your knowledge is good enough to make this kind of claim.

Basically, in your situation, what you need is a compelling example (e.g., a situation/use-case where your approach is either the only viable one, or where you can show with experiments or formal proofs that yours is better than others). If you can focus on that point, then you can probably resubmit it to the same journal, but it's possible that you will get some of the same reviewers you had the first time.

Another solution could be to publish the paper first as a technical report (e.g., on arXiv), try to submit it a short version without the proofs and technical details to a specialised workshop, where maybe you can get interesting feedback from the community (for instance, some recent related work), and then use that feedback to improve the paper.

  • @Noble: I'm thinking about it, as I dabble into revising the paper. I did want to see if there might be other answers. – Mike Dunlavey Sep 25 '12 at 17:36
  • Accepted, mostly because of the point that a new idea must be shown why it is advantageous, while conveying respect for prior art. – Mike Dunlavey Oct 01 '12 at 15:55
  • OK, I revised it along the lines you suggested and resubmitted it to the same journal. The editor was dubious of course, but he didn't refuse to give another shot. The trouble is, I always seem to pull my punches. I keep thinking afterwards of how I could have made my central point more crystal-clear. Oh, well. – Mike Dunlavey Oct 22 '12 at 12:51
  • @MikeDunlavey: Well, good luck with that new submission! But in that kind of situations, if you don't have any particular deadline you need to respect, then it might be worth asking some colleagues/contacts from academia for an informal review. That prevents your paper to be stuck for some months in an official reviewing process, and you have the opportunity to improve the exposition of your central point. –  Oct 22 '12 at 13:02