1

I have a work ready to publish with good results and proper mathematical arguments to justify it. I liked to try top conferences like ICML, IJCAI, CVPR, ICCV ... which are related to my field (machine learning and vision), but their deadlines are already passed or not reachable.

My supervisor has the idea of sending extended version of the paper (with extra mathematical proofs and experiments) to a top journal like JMLR or such.

But i'm not sure if it is a good idea, because it may take 6-8 months until i get the review and revise it and maybe 1 year until it gets published if they accept it, and my PhD will be finished before that!!

What if i send the paper (small version) to a medium conference which has a deadline in a month from now and also sending the extended version to a top JR? But i'm afraid by doing that i may undersell my work to medium quality expectations.

1- Is it allowed?

2- Is it a good plan?

  • 2
    1 - yes, 2 - yes. You first present something smaller on a conference (might call it preliminary results), and then further extended results, in a journal. It happens all the time (at least in physics/astrophysics), also the other way around. – corey979 Feb 8 '18 at 13:26
0

It is allowed, and it is a good plan.

More specifically, sometimes, conferences and journals will state, on their website, what proportion of the work you submit must be "novel." (How this is quantified is usually vague but it should give you an idea.) As long as you observe their rules you should be fine. Do keep in mind that it is likely for the same people to be reviewing papers for different conferences or journals, and if you end up sending insufficiently original work to multiple venues you may upset them.

Presenting research at a conference and submitting an expanded version to a journal is done very often in many fields. There are also many cases of journals inviting authors to contribute papers based on a previous conference presentation.

Regarding your more general question of "how to decide between submitting to a conference and a journal," I can offer the following strategy:

  1. Go to Google Scholar.
  2. Find the Metrics section.
  3. Go to the "top publications."
  4. Pick your category (e.g. Engineering & Computer Science) and possibly subcategory (e.g. Computer Vision & Pattern Recognition or Artificial Intelligence).
  5. Start from the top of the list and move down until you find a relevant venue with an appropriate deadline.
  • 5. Start from the top of the list and move down until you find a relevant venue with an appropriate deadline. I checked for one field related to my research interests (Discrete Maths) and I find out a quite bizarre ranking: Discrete Applied Maths (an average, average+ journal) ranked on the top of most prestigious journals such as JCT, Combinatorica, SIDMA etc. I do not trust this metric at all. – PsySp Feb 10 '18 at 11:47
  • @PsySp - To my knowledge these rankings are based purely on quantitative measures (citation count etc.). It may be that they don't reflect the true sentiment. It's also possible that for certain fields the data used by Google is of lesser quality. – mbaytas Feb 11 '18 at 12:36

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.