I am a 3rd year Ph.D. student. My field is computer science. I've worked on my topic for a long time, but finally, I found out that there are a lot of parameters correlated with publishing a paper. So I don't have hope for publishing a paper, since I should do it all on my own. What should I do now?

  • 1
    We need to know your field. We also need to know your goals - the correct advice is depends on whether you're aiming for a Harvard professorship or if you need a published paper because some university's regulations require one for you to be hired and don't really care what is in it. Commented Mar 31, 2019 at 0:02
  • My field is computer science and I'm working on the machine learning in network analysis. Actually besides every university rules or progress in my career, I need paper for my confidence. Because I read paper, have ideas, but l don't know why I can't distinguish valuable ideas from non-working and I also I think my ideas are obvious for everyone and doesn't have strong theory behind them.
    – user137927
    Commented Mar 31, 2019 at 7:35

2 Answers 2


Apply to lower level journals (not conferences, avoids the issue of travel). Just do some solid basic work to get your confidence up. Not "home runs" or "triples" but singles or bunts. Anything other than a strikeout.

Story: I had a girlfriend who was hired to sell huge enterprise level deals for a packaging company (not software, the boxes, but also box systems, services, etc.) She was hating life because she hadn't sold anything and didn't feel she was getting support to sell these huge deals and all the other salesmen would drop a price list and make little sales (not systems). I advised her to (counterintiutively) go ahead and sell some simple stuff like the Cro Magnon salesmen had advised her. The problem was that she basically had ZERO for a score and hadn't even learned how to use the systems (order entry to cash, credit checks for new customers, etc.) And her confidence was in the dumpster. She took my advice, felt WAY better after making a little sale and eventually got one enterprise deal done.

You need to get your confidence up. Just go to some decent journal. Not a fly by night predatory thing. But second tier decent, looking for datapoint papers. And don't try to claim you've invented hydrogen. Simple, simple, simple. And short ("least publishable unit").

P.s. When asking questions here, it helps if you tell us the field. There are huge differences by field. It's good if we/you can generalize as well (not requiring answers to be field specific), but if you give us the field as well, you will get some targeted answers from practitioners. I'm assuming it is computer science based on "stack exchange" and the comment about conferences, and I am not as familiar with that field, but many others here are.

  • I'm so thankful for your reply. Yes, my field is computer science. My adviser doesn't let me to submit my paper to lower conference or journal. He insists that unpublished paper is better than publishing in low conferences and journals. He wants us to work, but it makes me nervous that I can't publish in top conferences and high impact journals. Again thanks for your advice.
    – user137927
    Commented Mar 31, 2019 at 7:42
  • 2
    +1 I strongly disagree with OP's advisor. The first thing I try to get my own PhD students to do is just publish something. It doesn't matter what, and as long as the venue isn't a scam, it doesn't matter where. The confidence boost from having a publication outweighs the disadvantage (if any) of listing a weaker venue in your CV.
    – JeffE
    Commented Mar 31, 2019 at 14:18

Since your field is computer science, conferences are important. Much more so than in other fields. I don't know the academic culture of Iran, but in the wider world, your advisors permission isn't needed to publish anything anywhere. Also, in the wider world, degrees in CS are usually awarded for dissertations than for a set of publications, though this varies by country. In the US, you don't normally need publications to complete a degree in CS, though it is valuable if the ideas are worthy of publication.

I don't know what publishing without permission would do in your circumstance, so I'll assume the worst - that it would harm you. Otherwise, just do it. Keep improving the papers and keep submitting until you find success.

You don't seem to be without ideas. But I worry that you may be abandoning them too soon. That may be a problem, especially in a fast moving field. Timing is partly an issue in such fields. If you overly refine a paper before first submission then it may wind up too late in the queue to be considered. It might even be advantageous to submit something less refined, but with a solid core, so that you get in the publication process early enough to be considered.

Perhaps the following would help. Understand that your advisor is partly responsible for your success. Use him/her to sign off on the quality of your work and make suggestions to improve it. Suggest that you co-author a paper, putting the advisor's reputation on the line. You will probably learn something from this process, either how to write better papers or that your advisor is unsuitable. But both are valuable.

The fact that your friends have papers is partly serendipity. They were, perhaps, lucky. If they have the same advisor, then it is likely that there are good ideas floating around.

But, in general, find the shortest path to completion in your case and follow it. It might even involve changing advisors or institutions, though that is normally not a short path. But if you can finish with only a dissertation, then do that. If you can publish without permission, then do that. Etc. You don't need to stay in the current situation longer than necessary. Find a door and aim for it.

  • In the US, you don't normally need publications to complete a degree in CS — You don't??!
    – JeffE
    Commented Mar 31, 2019 at 14:20
  • @JeffE, you need a dissertation acceptable to the faculty. But publishing it is a different matter.
    – Buffy
    Commented Mar 31, 2019 at 14:22
  • Yes, I know the difference. Publication is not a formal requirement for a computer science PhD in the US, but (at least in all the departments I'm familiar with) it definitely is a de facto requirement.
    – JeffE
    Commented Mar 31, 2019 at 15:25
  • @Buffy Your answer is amazing. I can't submit a paper without my adviser permission, if I do, I might get fired and when I send a paper for him, it takes him a loooooong time to read and comment. You are exactly right. When I get rejected or have any idea, I abandon it so soon since I always think that this is an easy idea and after rejection I think it's unworthy to work more on it. I'm also too afraid that the field is moving and my idea perhaps is too old.
    – user137927
    Commented Mar 31, 2019 at 15:38
  • The main problem of working as team is that everybody is focused on his/her work and doesn't have time to help, so we are almost alone in papers. And since I'm alone I think testing every idea takes a long time to test and I'm always stressed that another group might work on the same idea. Again thanks for your comment. It is a good motivation for me.
    – user137927
    Commented Mar 31, 2019 at 15:38

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .