I have been working on a conference abstract recently.
Background for those who are unfamiliar with how presenting at conferences works:
1) You find a scientific conference you're interested in attending
2) You typically present either a poster or at an oral session, so you find a session that best matches your research area
3) You submit an abstract (and sometimes a short paper) to your session's chair(s) online
4) When your abstract gets accepted (there's no strict entry guidelines for giant national conferences, so likely it will get accepted), you book your plane tix and go
The problem:
Abstract/paper submission is done many many months in advance of the actual conference, especially for the huge conferences.
Why this makes sense:
There would be no time for session chairs to review abstracts at large conferences otherwise. There would be no time to publish and bind conference programs listing presenters either.
Why this doesn't make sense:
No one has their data pre-packaged and ready to go half a year in advance. The general consensus is that you write vague, related stuff, or package up your old stuff and sprinkle in some predictions and findings. Then you make the science happen.
I think most would agree this is sub-optimal as far as proper scientific approach.
The other general consensus is that no one reads your abstract anyway, so no one will ever hold you to the fact that you said you'd present on "Monkey interactions with banana-shaped robots" when in actuality you end up presenting on "Robot-monkey love: evolving interface programming." Hey, at least they both involve robots and monkeys. (The crowd might be angrier if they showed up to hear about robots and monkeys and then you surprise them with a talk on "Groundhog habitats in space")
So why can't we come up with anything better than this system?
Here's one idea: Let's submit a short abstract half a year in advance to make the program coordinators and printers happy. Then we submit a real short paper/poster one week in advance of the conference- to be submitted and available only electronically. And we shall ponder whether or not this piece is optional...
Incidentally, when I said no one reads your abstract, I should add that the 2 biggest things people look at when choosing talks:
1) Who's giving it? (Famous professors win. You don't have to be one- it counts if they're a co-author)
2) What's the title? (Relevance and sex-appeal, ie, a currently trendy topic wins)
5 weeks ago
I'm curious about the actual domain in Engineering that you're doing research in - if you feel comfortable with sharing that. I'm asking because I am doing a PhD in Computer Engineering and our story is very different.
ReplyDeleteWe submit to conferences 10-12 pages of ACM/IEEE format documents on a completed research project/idea. There is a committee that reviews the papers based on novelty, technical content, advance of the state-of-the-art, etc. Each committee member reviews X papers (some conferences have external reviewers as well); later on, the entire committee meets and decides the accepted papers. The acceptance rate varies from 12-25% (for top conferences). Once the paper is accepted, you have a month to prepare the "camera-ready" which is a polished version of the submitted paper. The "camera-ready" version gets published in the conference proceedings. 2-3 months after submitting the camera-ready, you go to the conference and present the exact paper that is published in the proceedings.
As a consequence of this procedure, in my domain, publishing in conferences is more important than publishing in journals. Journal publications are usually 2-3 conference publications + more data or small extensions.
I should probably add that I'm doing research at the border between Computer Engineering and Computer Science (i.e., my research domain is part of either or both departments, depending on the institution).
Very interesting indeed. Thank you for sharing a different approach to conference publishing. My guess is it makes for much higher quality conferences... Many of the large professional conferences I'm aware of are primarily used for networking and scouting out potential new talent. The faculty I've spoken with from the top MRUs don't think much of these conferences beyond this purpose.
ReplyDeleteSmaller, niche conferences do exist however where abstracts and submitted papers are highly scrutinized. These conferences have a cap on the number of attendees and are considered higher quality. Still, publishing in journals remains the key.
I did a 30 sec search for large professional conferences in Comp Sci and found a list that looked more specialized by large topical areas than the types of conference I'm talking about (where, say, ALL computer scientists would be gathered at one conference). I wonder if this is because Computer Science is one of the largest engineering disciplines? An analogy to my conference example would be an annual, national conference for IEEE (or some other large professional organization).