11

There have been a multitude of questions about the fact that in computer science and a couple adjacent fields, the standards conferences are held to is quite different than in most fields. CS conference papers in general are peer reviewed and count just as much as journal papers, the most prestigious venues in Theoretical CS are conferences, and it seems that CS poster presentations are assigned a higher credence than in most fields. These statements are very different than what you'd find in many other fields (political science, philosophy, physics, biology).

I work in an interdisciplinary field in industry adjacent to computer science, and some of my work is submitted to CS people and CS venues and other work is not. I don’t expect everyone reading my resume to know the differing standards, and it seems misleading to list, e.g., computer science and political science conferences next to each other.

I was thinking that on a CV/resume I might draw a distinction between “Peer Reviewed Conference Papers” and “Conference Papers,” but I was wondering if anyone had suggestions about how to best disambiguate this in contexts like writing a CV or grant application or whatever where you might not be able to just explain the difference and it matters that people understand.

  • 2
    Computer science conference presentations are not peer-reviewed. Computer science conference papers are peer-reviewed. – JeffE Jul 19 '18 at 13:22
  • 1
    @Buffy How would someone review the presentation? Someone can send an abstract that says one thing, but then the presenter actually says a subtly different thing during the talk... What happens then? – user9646 Jul 19 '18 at 13:38
  • 3
    @Buffy No, it isn’t hard to make a firm statement at all. Most CS conference papers are formally peer reviewed, in practice with the same level of scrutiny as journal papers. In many subfields (but sadly not mine), conference paper reviews are double-blind; I’ve never heard of double-blind journal reviews in CS. CS conferences that actually publish proceedings without review are rare, and largely considered to be scams. – JeffE Jul 19 '18 at 13:38
  • 1
    @JeffE, sorry I wasn't more clear. I was referring to poster sessions only. I think both you and the OP made statements that were too strong. Yes, CS conference papers are properly reviewed and for some poster sessions the submissions get some review. – Buffy Jul 19 '18 at 14:21
  • 1
    @Buffy Whether a result is presented at a poster session is not relevant; the relevant question is whether the presentation is attached to a paper in the proceedings. At some conferences, some accepted papers are presented by talks and others in poster sessions; in those conferences, both classes of papers are stringently reviewed. At other conferences, poster sessions are not associated with proceedings papers; those posters sometimes go through a light sanity-check, but nothing approaching full peer review, because they aren't attached to a publication. – JeffE Jul 20 '18 at 1:41
4

Having been on a couple hiring panels for people like you - and indeed being in the same situation (for a long time my statement was that my two most prestigious publications were each recognized as a thing by exactly half of the people reading my CV), a couple suggestions:

  • Cover letters. It might be appropriate, in a cover letter, to call attention to conferences that are particularly prestigious and state that.
  • Reference letters. If you have authors on the CS side of things, ask them to push that expressly. Talk about how impressive your work is and what a big deal those conferences are. This is what I asked my letter writers to do. Regardless of field expectations, it's important to have your letter writers put your work in context.
  • If possible, try and find out who is reading things, and see if you need to do this kind of advocacy. For example, for an interdisciplinary hiring panel we had, we expressly included someone from each discipline so that they could provide context - because even though I know CS uses conferences as their major venue, I still can't distinguish between a great, good, and just okay conference.

I think it's appropriate to make the distinction in your CV as well - like the difference between submitted and invited talks, they are somewhat different animals.

2

I think you are wise to make a distinction. Computer Science and some related subfields use conferences as the major way to disseminate new results with journals more or less reserved for more settled ideas.

I can only speak for ACM (i.e. not IEEE), but many of the conferences are extremely important, SIGPLAN has several such and the SIGGRAPH conference usually introduces remarkable results. The Agile Software Development community holds a number of conferences with peer-reviewed work, but has no journal. The Software Patterns community works on a completely different model, publishes the work of conference-workshops, but again, no journal.

Part (maybe all) of the reason for this is the young (~75 years or a bit more: the Turing Machine was described in 1936) age of the discipline. Much of the development of the discipline has occurred in the internet age with frequent if informal communication.

Political Science, of course is a lot older (Sun Tzu, Machiavelli, ...) and more developed.

How to do it is another matter. If your work falls into the two disciplines separately it is easier - just have two sections in your CV. Different audiences will focus on one or the other and will likely know the expectations of the field they are interested in.

But a lot harder if your work is integrative of the two disciplines. It will be harder for each audience to appreciate the difference (A footnote that "Buffy sez" won't get you there, of course). I don't know of a reliable and official source for the things I've said here to which you could point, though there is probably some documentation of it "The importance of conferences to CS", say.

However, I think the issue is more important for a new member of the profession, rather than an established member as I suspect is the case of the OP. Your general reputation will be clear to people with no more than an introductory statement (among the rest) that you have published in appropriate and (respected) venues to the disparate fields.


With respect to grant applications, I think you should actually have more flexibility to inform reviewers than in a CV. If the granting agency is more concerned with one field of the two or more you are covering, then they will be mostly concerned with your references in that field and wouldn't know much about the others in any case.

But you could give as part of your background, I hope, in most cases, an explanation of the importance of conferences in CS re journals. It would only take a short paragraph and a reference or two. However, I don't think it would really be critical unless the grant was primarily NOT CS but the reviewers had to understand how CS actually works in practice. But I think in such a case (never actually having danced on the cusp, here) that they are more interested in the "other" stuff and see the CS part as primarily methodology. But you need to be able to back up your methodology in any case. The CS folk will "get it" and the others will think the "other stuff" is the most important anyway.

I find it hard to believe that a political-science grounded grant would fail for reviewers thinking you weren't a mover-shaker in CS. What I mean by this is as follows. If I'm a political-science guru reviewing your grant application to explore some deep topic, I'm totally focused on my own field and on what I think is your potential contributions to it. The fact that your methodology uses, say, machine learning, will be foreign to me and it will be hard for me to judge in any case. But I'll be intrigued. However, if I think you are primarily interested in CS and are naive about political-science, I vote no. But if you are clearly well published in the field I'm interested in, and your arguments are good I'll vote yes without a deep dive into your position in CS, though I will need "sufficient assurance" there. The one overbalances the other. I suspect I'd be intrigued by a new approach to things I consider important.

However, if I'm an AI guru instead, I understand the quality of your CS publications and will focus on that. I'll see the political-science stuff as an intriguing application of what I consider important. But here, I won't know about the quality of the political-science journals you are published in without asking someone.

The importance of journals v conferences is asymmetric, but the problem you face is, itself, asymmetric in a matching way. Pretty sure, anyway.

  • 1
    The sort of document mentioned does sound like it would be of use to a lot of young researchers. Something along the lines of the "culture statement" from the AMS about author order in math. – Tobias Kildetoft Jul 19 '18 at 17:15
  • [Here](www.ams.org/profession/leaders/culture/CultureStatement04.pdf) is a link to the 2014 culture statement @TobiasKildetoft mentions if anyone is interested. I agree that it would be great if ACM, IEEE, or a similar group was to make one for computer science. – Stella Biderman Jul 19 '18 at 18:21
  • Thanks a lot for this answer, it feels validating even if it doesn't quite answer my question (as it lacks any particular advice or response to my suggestion). I am a young researcher and my work involves applying computation to solve problems in political science and related fields, so it is unfortunately in the interdisciplinary realm you describe. – Stella Biderman Jul 19 '18 at 18:24
  • 1
    @StellaBiderman, reading it again, it sounds like "Grant application intrigue". Maybe it should be a doctoral level course. – Buffy Jul 20 '18 at 12:52
  • 1
    Perhaps you’re looking for somthing like this? cra.org/resources/best-practice-memos/… – JeffE Jul 20 '18 at 13:48

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.