24

My paper got accepted normally (e.g., with neutral to positive reviewer rating) to a conference.

I have noticed that, I always get to present (oral or poster) on the last days of the conference, and worse, at the last session.

Predictably, I did not get a large audience, and got fewer questions/useful feedback compared to papers presented on the first, and second day of the conference.

Does that indicate (low) perception of my paper by the conference organizer?

It is almost as if they want to conceal my paper. The conferences are ranked typical B-level with ca. 50% acceptance rate.

What can I do?

15
  • 9
    Are you saying that this has happened multiple times, at different conferences? Are you suggesting that the organizers of the different conferences have conspired against you? Commented Jun 6, 2023 at 3:27
  • 10
    Is your name or surname starting with X/Y/Z? it may be a very unlucky crossfeedback between your session of choice and your name.
    – EarlGrey
    Commented Jun 6, 2023 at 7:34
  • 8
    How many instances are we talking about? What is a B-level conference? What field are you in? Commented Jun 6, 2023 at 8:12
  • 20
    It may be nothing to do with the perceived quality of the paper, but more to do with the perceived level of interest in the subject matter, or of your personal "brand recognition". FWIW I once travelled half way across the world to a conference to find my slot moved to 8am on the last day after the conference dinner, and I had an audience of 7. Commented Jun 6, 2023 at 9:02
  • 4
    No, I did not mean "conceal" in conspiratorial way. More in the sense I suspect they think the paper is of low relevant, or perhaps quality (ouch!) so it's only there for meeting quota. I rather not be specific about the conference name, but I have counted 4 instances so far (one of them put me on the last day, but not last slot, to be fair). The field is computer vision. My name is not on the last alphabet no. @Michael Kay, I'm just a second year phd student, so obviously I have no "personal brand" to speak of.
    – KitKat
    Commented Jun 6, 2023 at 13:48

5 Answers 5

29

Your question has just made me doubt any conference I have presented at - I have been in the last session so many, many times and I always considered it just bad luck (getting a talk in the first place was the exiting bit).

While some of the other answers have already given you a nice perspective regarding the considerations that many conference organizers may have had when they were putting together the schedule, let me also offer my perspective, having now organized multiple different conferences and symposia, big and small: in none of these cases have we put together the schedule based on whether we considered talks selected from abstracts to have lower or higher impact or significance. As an organizer, I want my last session to be as strong as the first, if possible, because nobody wants an empty room towards the end of the conference (what has happened therefore: putting an awesome keynote speaker or exciting invited talk at the end to make sure nobody heads home).

Even when picking best speaker awards I've often been in last minute hushed discussions as we considered a talk from one of the last sections without a coffee break in between the session and the actual award ceremony.

I do agree that it sucks to be on the last day because you lack the free conference wide exposure you might get by presenting on the first day, but there are other ways to put yourself out there: sit with people you don't know at breakfast/lunch/dinner, be active and introduce yourself to people at poster sessions etc. etc.

So, don't get worked up over it, don't read too much into it and just focus on the contents of your work and on genuinely getting to know people in your field.

4
  • 4
    Thank you for the encouraging explanation. I did try to speak to people with similar sub-topic in the first/second day of the conference (i.e., after their oral or during their poster session). Sometimes I casually bringing up my works, and if I'm lucky I get the feedback that I need. A lot of them were already gone by the third day! So to whoever shares the same fate as me, this is the least you can do. :)
    – KitKat
    Commented Jun 6, 2023 at 13:58
  • 10
    "none of these cases have we put together the schedule based on talks we considered to have lower or higher impact" seems to contradict with "the opposite has happened: ..."
    – Dan M.
    Commented Jun 6, 2023 at 14:06
  • 2
    +1 for the clustering by topic, and the odds and ends/not enough talks to fill one whole session. This has been my experience as well. Sometimes to the detriment of the clustered sessions as no 'new' people would attend those. Commented Jun 6, 2023 at 18:13
  • @DanM. well spotted, edited for clarity.
    – BioBrains
    Commented Jun 8, 2023 at 5:28
11

How do conferences group papers?

(Based on my experience and therefore subjective)

Usually, the programming committee has a meeting to discuss referee reports. On rare occasion, last-minute reviews are being done by PC members to balance the fact that reviewers did not respond timely or in order to understand the review reports better. Before the meeting, the PC chair might divide the submissions into three groups: definitely accept, where the PC is expected to rubber-stamp this decision, definitely do not accept, for instance, because none of the three reviewers voted for a straight accept, or as desk-rejects, and the papers that ought to be discussed. In parallel, the number of papers to be selected is determined.

When the work is done, the Chair with or without help groups papers together by threads. Some papers will cluster naturally, and sometimes they will not, leading to "Odds and Ends" sessions. After the sessions have been determined, they get assigned to dates in the calendar. Even if there is no Best Paper award, the papers that according to the PC and the reviewers are or were candidates, have their session placed in the prime time, usually a session after the key-note talk. Everything else gets a spot assigned usually without much deliberation. Odds-and-ends sessions without a strong paper and without strong cohesion tend to get assigned last session in a day.

Now, big conferences like VLDB and those where the tracks are more independently organized work differently.

TO ANSWER YOUR QUESTION:

It is unlikely that someone is out to get you, because the decision making process is too distributed for that to work easily. It might be a sequence of bad luck (statistically slumps are likely to occur) or it might be that your work does not cluster well with other work.

Unfortunately, the conference method of disseminating information has draw-backs, and you are suffering from one. If you look at the history, conferences started out much more as an exchange of ideas before the developed into the main dissemination method of Computer Science.

If conferences accept your papers, someone thinks your work is worth-while. However, even if you end up presenting in a prime spot, your audience tends to be already overwhelmed by the program and you might not get a substantially better reception.

2
  • 2
    No, no, no. I didn't mean "get me" in such conspiratorial way. No one care that much about some unknown phd student, haha. More like, in your words my paper was more than often perceived as belonging to "Odds-and-ends session". If that's true, I'm trying to improve my works and would like to know where to start.
    – KitKat
    Commented Jun 6, 2023 at 14:03
  • 4
    Working on stuff that is a bit out of the main stream is actually cool. So, with "odds-and-ends" I do not mean bad or inferior quality or "just squeaked by". I remember one PC meeting where the PC was quite enthusiastic about accepting a paper because it was a bit different, even though it fit the call-for-papers. Commented Jun 6, 2023 at 16:20
9

I very much doubt that the organizers want to conceal your paper. If they thought that they would have rejected it.

I suspect that although it cleared the cutoff point for acceptance, the organizers thought it less interesting or more specialized than papers scheduled earlier. You might be able to infer that by looking at the rest of the schedule.

As for what to do? Approach other participants whose work touches yours. You may learn something; that may encourage them to come to your talk. Try harder to connect your work to questions people at that conference care about.

That said, congratulations on the acceptances.

1
  • 2
    I hope so, by conceal I mean in the "hope no one noticed I put this paper here" sense. It is true that my subtopic is more relatively specialized. Yes, I have been approaching people on the first and second days, because often by the third day they were gone!
    – KitKat
    Commented Jun 6, 2023 at 14:07
7

It seems very unlikely that there is a conspiracy to bury your results at the end of conferences. I think there are a few possibilities:

  1. Your work was considered less interesting and therefore not worth a "prime-time" spot. It's impossible for anyone here to determine this and, honestly, its not worth thinking about.
  2. The conference organized posters/presentations based on some other factor like order of submission, name of author, topic, etc... Anything could have landed your presentation at the end - someone has to be last and it might not have much to do with quality.
  3. Related to 1, are you sure your presentation was really best suited to a particular conference? Its always possible that your work is actually quite good but not 100% on topic (or of broad interest).
  4. Bad luck. Unless you have a trend spanning dozens of conferences, maybe there is nothing here. We tend to focus on extreme outcomes and you're more likely to notice when you are at the end (especially since you think this is a trend) than when you are stuck somewhere in the middle.

I don't actually think there is anything for you to do. It's not even clear that there is something wrong in the first place. You probably can't go wrong with @Ethan Bolker's suggestions. But really, just focus on producing quality work and identifying conferences which are best suited to sharing that work.

1
  • 2
    Ah again, I did not mean "someone is out to bury my works". More along the lines "well, I need to fill some sessions, this paper got peer-review acceptance, but is a bit dubious, if I put it here, I bet no one will notice". #3 is plausible reason for some of the conferences I attend, but definitely not all. Yeah, I will try my best to focus on the content.
    – KitKat
    Commented Jun 6, 2023 at 14:12
7

Mostly, my experience agrees with the existing answers by BioBrains and Thomas Schwarz — schedules are decided by a variety of concerns, with “quality” far down on the list if it’s considered at all.

The one common exception, in my experience, is that committees often aim to put strong talks in certain key time-slots — speakers we expect to give a compelling talk and/or attract a large audience (which mostly coincide). These slots typically include each day’s opening talk, especially the first day, but also often each closing talk as well, especially the last day’s, to make sure the conference ends on a strong note, and also to motivate participants to stick around to the end. An unpopular schedule on the last afternoon is a recipe for audience attrition.

So a few of the strongest (maybe c.10% of all talks) often have their quality taken into account in their scheduling; but for the placement of the rest, quality is essentially irrelevant to the schedule. In particular, I’ve never known a committee speak in terms of trying to “bury” bad talks, or consistently put weaker talks into later sessions; so I wouldn’t draw any bad conclusions from the scheduling you describe.

You must log in to answer this question.

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