Friday, February 11, 2011

PROS and CONS of being on a Program Committee



What are the PROS and CONS of being on a program committee?
  1. PRO: Looks good on your resume. Is this true at your school? This PRO may be more relevant for untenured and un-full-prof people. (My spellcheck wanted me to use unturned instead of untenured.)
  2. PRO: You get to see what people are working on. This may give you ideas of what to work on or of what papers to look at. You absolutely cannot use the ideas you see until they are in the public domain; however you can learn things and look up some past work.
  3. PRO: If there is an in-person meeting then you get to meet some new people. (If there is NOT an in-person meeting then its not the same.)
  4. PRO: You get to see how the process works which may help you when you submit later. Or it may help you decide not to submit.
  5. CON: It takes a lot of time.
  6. CON: You may have to read papers that you don't care about. (This could be good for you.)
  7. CON: You may have to read bad papers. Not clear though- if you can tell its bad early on then you can skim the rest.
  8. PRO AND CON: IF you also goto the conference then (a) you will understand more talks (b) you will be bored at more talks. I tend towards (a) but others tend towards (b).
  9. PRO AND CON: You get to help decide what papers get in. But you may see things not go the way you think they should.
I'm sure there are more PROS, CONS, and THOUGHTS- please share them!

18 comments:

  1. I really think 6 is a PRO not a CON. Reading papers outside your area is a good way to broaden yourself.

    Also, it actually is much more painful to read mediocre papers than to read bad papers. There is unfortunately some glee in lampooning really bad papers, or finding actual mistakes. At least you feel like the process will terminate, and you're doing some good. The worst is slogging through technically correct but uninteresting work to verify it, or trying to parse ambiguously written work, knowing that it is unlikely to get accepted anyway
    (but will appear in a less prestigious conference, or keep on getting submitted....)

    Russell Impagliazzo

    ReplyDelete
  2. "Not clear though- if you can tell its bad early on then you can skim the rest."

    I find you have to read the paper anyway so that you can clearly say why it should not get in.

    ReplyDelete
  3. Does this also apply to referees who do only three or four papers?

    ReplyDelete
  4. I think number 8 is a big pro. The best talks are ones you've heard before :-). And 6 too. Reading papers outside your immediate area is hard work, but important to do, and this forces you to do it.

    It is, though, a *lot* of work - you need to devote a solid month of your life (or 50% time over 2 months) to it.

    ReplyDelete
  5. what about those program committee members that REJECT a work, rush to write a similar paper (sometimes better than the original author) and submit it to another conference??

    please, tell me that that doesnt exist, please, I want to laugh today...

    ReplyDelete
  6. @Russell "Technically correct but uninteresting and unlikely to be accepted." Guess that's the reason of a recent rejection, although I can't understand why it is not interesting. There are definitely a large amount of papers accepted but not so interesting to me.

    ReplyDelete
  7. Are there cases where someone on a program committee rejects a paper and then rushes out to write there own version.
    YES

    Is it common in TCS?
    NO

    Is it reprehensible?
    YES

    How many credible rumors of such have
    I heard in my 25 years as a theorist?
    1

    ReplyDelete
  8. @Russell "Technically correct but uninteresting and unlikely to be accepted." Guess that's the reason of a recent rejection, although I can't understand why it is not interesting. There are definitely a large amount of papers accepted but not so interesting to me.


    How about rejection due to "interesting results on important problems but uses techniques from earlier paper"?

    ReplyDelete
  9. A further CON of being on a PC is not being allowed to submit papers to the conference (or being restricted to submitting one paper, depending on the community). Although this is completely fair, an advisor on a PC can have a non-negligible impact on his / her students' career.

    ReplyDelete
  10. How many credible rumors of such have I heard in my 25 years as a theorist?

    I had two good papers rejected one in theory, one outside TCS, only to see highly overlapping papers by members of the program committee rushed into publication soon after. Since then I've heard at least two other people complain about the theory PC member above. Is it suspicious? yes. Am I positive about it? no.

    ReplyDelete
  11. Boring papers might get rejected, but I didn't mean to imply that all correct papers that get rejected are boring. Plenty of very good papers get rejected. Sometimes they get rejected despite the protests of one or two PC members.

    On the topic of PC poaching, there is another possible explanation when a PC member publishes a result related to a rejected paper soon thereafter. When I'm on a PC, I cannot submit my papers to that conference. So often this means that I'm delaying publication of some results, which may explain why it looks like PC members are ``rushing work into publication'' for the next conference.

    If there's a clear overlap between what I've done and what is in a submitted paper, I'll declare a COI and abstain from reviewing that paper. The other members of the PC might not like the subject, even if I do. After the decisions are made, I will want to publish. Ironically, if that other paper was accepted, I can notify the authors of my related work.
    But if it was rejected, I am supposed to pretend I don't know it exists, unless I can find it some other way (author's website, arxiv, ECCC). There are work arounds, but it gets hairy.

    Russell

    ReplyDelete
  12. Russell: The PC member also has the option of posting on the arXiv.

    ReplyDelete
  13. I would add that is now difficult to separate which papers were good and not accepted (and so I shouldn't talk about them) from those that were accepted without google handy. All that wouldn't matter if people would just post to arXiv by default ...

    More here.

    ReplyDelete
  14. "So often this means that I'm delaying publication of some results, which may explain why it looks like PC members are ``rushing work into publication'' for the next conference. ... But if it was rejected, I am supposed to pretend I don't know it exists, unless I can find it some other way (author's website, arxiv, ECCC)."

    Pretending the paper does not exist and going ahead and publishing one's own paper on the same subject without contacting the authors seems highly unethical!

    The only ethical things to do in this situation are either:
    1. Give up the research, or at least wait for the other paper. After all, the PC member knows that he doesn't have priority. (A special case is if the PC member has submitted the research to a different conference at the same time. If it was really ready, he should have posted it to the arXiv, but older PC members still figuring out the Internet might not know this.)
    2. Contact the authors. Yes, it contradicts what a PC member is supposed to pretend, but compared to publishing a paper on the same subject without contacting them, this option is the lesser of two evils!

    ReplyDelete
  15. For me the biggest PRO of serving on a PC was not so much seeing what people were working on or meeting people and getting exposure on the PC.

    The big thing I learned (in face-to-face meetings) was how leaders in the field valued and thought about TCS research at a meta-level. I had been exposed to this in a limited way in my time as a grad student and postdoc but this was on a completely different scale, in an environment where the tradeoffs brought out this meta-reasoning front and center.

    I had always assumed that the quality of the theorems and strength of results were what should matter above all. I learned to appreciate that there could be higher values. Number one among these was the value of novelty. Not in the narrow technical sense of novel results but more in the sense of new ways of looking at things. Related to this was the value of bringing new tools into the field independent of the problems being solved with them.

    Another value that I had given too little weight was the value of elegance and simplicity. I had assumed that hard technical work was better but simplicity and elegance are evidence of greater understanding.

    Of course what you learn when you serve may be different...

    ReplyDelete
  16. I second Paul Beame's comment.

    ReplyDelete
  17. Anonymous but not Brainless8:01 PM, February 12, 2011

    Some random anonymous kook suggests that if you are on a PC and there is a junk paper submitted that happens to be on a same or similar topic as yours that you should ditch your research? I didn't realize this was a comedy blog.

    ReplyDelete
  18. not junk, anonymous genius, but serious papers with interesting results that helped you achieve your objectives.
    who's the real kook?

    ReplyDelete