7

My paper is accepted in an A-rank conference in computer science as a short paper. I need to reduce it to 6-pages instead of 10-pages. All the reviewers said the paper has all the required elements to be a decent paper, but the presentation and problem introduction should have been better.

Also if i have to cut 4 pages, I need to reduce the references and the propositions and the proofs, and some of the related works!

So I'm not sure if it is better to withdraw it from this conference and submit the revised full version to another one? Or i should take the opportunity of having my paper in this top conference and try to submit the extended version to a journal paper?

BTW, i'm almost at the end of my PhD study and have not yet published any paper in a top-tier conference, and this work is one of the main results of my work (however there might be one another work to be submitted to another A-rank conference).

  • 4
    It is more valuable than not having one, certainly. – Buffy Aug 21 '18 at 12:37
  • 2
    Defer all of the proofs (even the short ones) to a (10 page) technical report (which you reference), drop some references, compress the main body, and publish a 6 page paper. – user2768 Aug 21 '18 at 12:44
  • @user2768: good advice. How about also submitting the extended version as a journal paper? of course with additional experiments and more analysis of the results and the method's characteristics. – Bob Aug 21 '18 at 12:57
  • @Bob Whether a journal article is merited is unclear, because there's insufficient information. You need to determine whether the four extra pages in the 10-page version contain sufficient novelty to merit a journal article or whether those four pages can be further supplemented with "additional experiments and more analysis of the results and the method's characteristics" to merit such an article. – user2768 Aug 21 '18 at 14:05
  • 1
    @OlegLobachev and cite it in the short version for the omitted proofs! – Stella Biderman Aug 21 '18 at 23:17
5

I'm wondering how valuable is to have a short paper in a top conference?!

That depends a bit on what you define as "valuable". As Buffy says, having a presentation and presence at a top conference is certainly very valuable in terms of visibility in the community and opportunity to make connections (never underestimate the importance of those!).

However, in my experience, as a line in your CV, everything that's not a full paper counts for very little, even if it's a short paper at a top conference. In other words, if that is intended to be the main paper of your PhD and you intend to go onto the job market with this work, you need to get it accepted as a full paper. If it's just a side project, or just one in a longer string of papers that together form your PhD, leaving it as a short paper and taking the opportunity to attend the conference and get your name out there may be worth it.

I should also say that, assuming your reviewers are right and all it needs for this work to become excellent is some presentation-level cleanup, it feels like a pity to not do this small additional work and try again at the next deadline for a top venue, rather than downgrading it to a short paper.

  • Seconded. The advice I have always been given is that short papers usually aren't worth the time and effort. Do what you can to get the work up to par for a full paper, unless you just want the visibility from presenting. – Austin Henley Aug 21 '18 at 16:55
  • @xLeitix: I want to count on this work in my CV as it is one of the main results of my PhD study (is about to be finished), and i do not want to waste it just for being in an A-rank conference. I can be in those conferences even by submitting posters to their workshops! – Bob Aug 21 '18 at 18:55
  • @xLeitix: Maybe you are right about the big gap between the value of full papers and short papers in a top conference, however, short papers are not a chunk of rubish rejected on-the-progress works! After the accepted full-papers, they were the most top-ranked submitted full-papers. The acceptance rate is 8.8% (full-papers) and 11.4% (short-papers)! – Bob Aug 23 '18 at 8:41
  • @Bob I did not say "rubbish", but I will stand by "in-progress". Fair or not, this is what many people think when they see a short paper (often because this is what short papers tracks are nominally for, at least in the conferences that I am involved with). – xLeitix Aug 23 '18 at 9:11
  • @Bob Also, acceptance rates can be very deceiving. You need to know what is submitted to a conference or track to say whether a specific acceptance rate is good or bad. I know one notorious conference which has always had an acceptance rate of about 16%, and the papers are generally not great. Contrary, FSE, an A* conference, usually has acceptance rates of 20+%. The difference is that most people don't even try to submit horrendous papers to FSE, while everybody seems to submit every crap they have to the other conference. – xLeitix Aug 23 '18 at 9:13
3

I don't know where you are in the profession. I'll assume you are relatively new. I might answer differently to a full professor. But what any presence in a good conference gives you is visibility and the opportunity to meet and interact with others who can help you along the way.

Granted that reducing a paper by 40% is quite difficult, but if you can do it, tightening up arguments, for example, then it may actually turn out to be better. Perhaps you have other advice from the referees as well. Perhaps it is possible to leave some parts out for a later, longer, paper.

But if you withdraw a relatively certain presence for a possible future one, you may actually be harming your career.

If you were limited to one paper in your whole career it would be a different situation. But if this is early work, I don't think there is any downside to this other than the time it takes. The work that goes in to it may even give you fresh ideas for the future.

  • I'm in the last year of my PhD, and i do not want to waste this work if i know it doesn't count that much in my CV! At the worst case, i can publish it as a full paper in a B+ conference! – Bob Aug 21 '18 at 19:03
  • But going through the he... sorry, review rounds of the B+ conference would also take some time! I am for shortening, if this matters. – Oleg Lobachev Aug 21 '18 at 19:24
0

I'm not in computer science, but this is what the situation sounds like to an outsider: from this,

I need to reduce it to 6-pages instead of 10-pages. All the reviewers said the paper has all the required elements to be a decent paper, but the presentation and problem introduction should have been better.

it sounds like the reviewers are pretty confident that with the results in your current manuscript you should be able to write a single paper that is equally good and equally complete within those six pages, and without needing to farm out any substantial parts of your proofs to external (appendices / arXiv / journal paper / weaker conference paper) documents.

This will obviously sound like it's somewhere between undoably hard and truly impossible to you, but you should give due consideration to the possibility that your abilities to write concisely and compactly still need to improve and that this is what the reviewers are asking you to do.

(This assumes, of course, that the only difference in what the reviewers are asking for is the length. If they are indeed proposing to drop you from 'full paper' to some other category, in CS-specific ways which are not evident in the question text and which would impact your CV beyond just the number of pages, then the situation obviously changes.)

And, all of that said: why are you having this conversation with strangers in the internet instead of with your PhD supervisor?

  • only one of the reviewers (out of three) voted for short-paper, and he was from a different domain of expertise than my work as he made some general negative comments without going into any detail! So that does not mean my paper suits a short-version in the first place! – Bob Aug 23 '18 at 8:37

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.