8

I am a PhD student in computer science, working in theory. I was recently reading a research paper, which was submitted to a tier-3 conference last year. It gave me some idea (which may not work), which is a kind of incremental, but yields a significant improvement of the run time.

The only problem I am facing right now is where I should submit this research paper. Although idea is not big as per me, my colleagues are saying it is significant and I do not realise as I have been working on this for three months. This is not related to my PhD project. I asked my supervisor and he said it seems okay to him.

Question: Is it possible that an incremental work on research paper may get selected for a top conference? I have hard time deciding to which conference I should submit. I have little experience as I have already published two research papers but that does not seem to help much.

Wrzlprmft
  • 61,194
  • 18
  • 189
  • 288
ffffref54
  • 597
  • 7
  • 12

2 Answers2

10

It's possible, in the sense that there's a non-zero chance of acceptance. A couple of factors strongly contribute to this chance: How well can you explain the significance of the problem? How big is the novelty delta, for example, are there some entirely new concepts? How big is the runtime improvement? How rigorous is the evaluation?

In essence, you need to write a very strong paper, which is naturally easier for some contributions than for others.

lighthouse keeper
  • 25,771
  • 3
  • 65
  • 117
  • 2
    As I spoken to few colleagues it seems to them that significance of the problem is high, Idea is good. – ffffref54 Mar 28 '18 at 11:27
  • 5
    @Complexity My point is not that your problem must be significant (it certainly has to be), but that your paper also needs to do a particularly good job to communicate the significance. Especially when the problem was not considered at top conferences before, you have to be careful to get this right. – lighthouse keeper Mar 28 '18 at 11:47
4

If the work is significant and a good contribution, then it should get selected to a related good conference. The selection also matters if the conference you are submitting is related to the area you have contributed.

I ask my supervisor he said it seems ok to him

What do you mean by ok? If he agreed to submit to the good conference then go ahead and give a try.

Nothing wrong in giving a try to submit it to a top conference and if you are unlucky then submit it to a lower grade conference.

Agaz Wani
  • 349
  • 2
  • 5
  • Result seems ok as he has not gone through entire work he said he can't say more. – ffffref54 Mar 28 '18 at 07:36
  • 5
    I am afraid of this word "OK", sometimes it means nothing. So you only can tell what kind of contribution you have made (have you modified the algorithm, you have optimized some code etc.) – Agaz Wani Mar 28 '18 at 08:25
  • 1
    @Complexity - Is there a professor you can get to read your paper and (ideally) go through your work? It seems like your supervisor ought to make time to look at this before your deadline, if you ask early enough. – cactus_pardner Mar 28 '18 at 09:32
  • 1
    @Aaghaz Hussain I have spoken to post doc's – ffffref54 Mar 28 '18 at 10:47
  • @Complexity if you are confident enough, go ahead with the submission (Plan A: submit to the top conference. Plan B: If rejected, refine and submit to another ). In case of plan B also failed, refine your manuscript and repeat B. – Agaz Wani Mar 28 '18 at 11:15
  • 2
    @Complexity "OK" doesn't help us much. "Your paper is oooohkaaaaay," means it isn't bad but there's nothing really positive to say about it. "OK -- submit it!" means "Submitting it is a good idea", which suggests it's pretty good and has reasonable prospects. – David Richerby Mar 28 '18 at 12:55