Why don't North Americans (US+CA) publish work at the
MODELS conference? The conference is often outside the US, but even when it is located in the US (Denver), North Americans don't publish there in great numbers. Even then, there is a higher percentage of North American papers at ECOOP, which is never located in North America. We don't like AOSD very much either.
| Conference | Location | North American papers |
| Models 12 | Austria | 16% |
| Models 09 | US | 21% |
| AOSD 12 | Germany | 22% |
| ECOOP 11 | UK | 35% |
| OOPSLA 11 | US | 54% |
Of course, maybe they do submit papers but they just aren't accepted :-)
Update: The numbers for USA are even worse:
| Conference | Location | US papers |
| Models 12 | Austria | 6% |
| Models 09 | US | 12% |
| AOSD 12 | Germany | 22% |
| ECOOP 11 | UK | 31% |
| OOPSLA 11 | US | 52% |
To be fair, you need to show the percentages of papers submitted, not published. AOSD is/was a particularly insular group that went into some dark group-think times that shunned those "out of the loop".
ReplyDeleteMuch to the damage of the health of the conference, their good ideas went to ICSE/FSE instead. And then the cycle continues, with fewer and fewer submissions from great researchers who don't want to waste their time on insular PCs.
Just FYI, MODELS is generally held every other year in N America. The only exception is when it goes outside Eur/N America: eg to New Zealand in 2011. Modelling in academia has always been somewhat European; ironically, industry modelling is mainly american
ReplyDeleteEuropean funding sources are also more supportive of modeling work, with large programs shared across multiple nations. It is challenging to get funding at times (at least in the US) for modeling work, which then slows the pipeline for potential publications.
ReplyDeleteI've never been to MODELs so it is hard for me to know. Neil Ernst @neilernst tweeted that the same applies to CAiSE: I find there is only one US author in all the papers at CAiSE 12 in Poland.
ReplyDeleteThe comments give some good ideas, but none have really answered my question. The funding situation may be a consequence of some underlying beliefs, not a cause. Perhaps some more detailed analysis of other conferences might help, or interviews of participants?
One factor is that the US has a bias toward publishing in top conferences (OOPSLA, POPL, PLDI, ECOOP, ICSE, etc) and that second-tier conferences are not counted for nearly as much for promotion and hiring. My sense is that Europe does not have as strong a bias in this direction. I'm not sure if this applies here, but it might be a factor.
A few comments:
ReplyDelete- MODELS will come back to the US in 2013 (October) and will be hosted in Miami. Hopefully we can get more North American submissions.
- MODELS 2012 has an NSF grant to support Doctoral Students from the US to travel to Austria.
- MODELS acceptance rates are recently in the 20% range, with one edition at 15-18% acceptance very recently. I am not sure I would call MODELS a premier conference, but perhaps not fair to also call it second tier.
Don't measure conference quality by acceptance rates. It means either the conference is selective or is a hot topic, but nothing more. Quality is difficult to measure, but it certainly should include number of citations, and where these citations come from. We need a "google paper-rank and conference-rank" to help us figure this out :-)
ReplyDeleteThere are definitely continental (and even national) characteristics in research fields. Systems work is surely more a US thing than European; whereas program calculation is the other way round. Within functional programming, US is strict and mostly untyped, EU is lazy and nearly all typed. Similarly, there are characteristics of particular departments. Need there be a "why"? Perhaps it's just a ripple arising out of a historical accident.
ReplyDeleteI don't buy the "top conferences" argument. Not all US departments are top-tier; presumably second-tier departments do second-tier work which appears in second tier conferences. (But I have to say that the US tenure system looks brutal from the outside!)
Not just North America, there are also few papers in MODELS from UK, Australia, Japan, and China. Actually, I'm afraid a very large proportion of papers are from German-speaking countries (including Northern Europe) and French-speaking ones (including the French-speaking part of Canada). Then a few from Italy and Spain. So there seems to be a geographical factor :D
ReplyDeleteHi Will,
ReplyDeleteThis may be of interest (thanks for pointing out by Shriram Krishnamurthi):
Best Papers vs. Top Cited Papers in Computer Science (since 1996)
http://arnetminer.org/conferencebestpapers;jsessionid=E0A30E1D0F6EBD1F713AB5C29B79D8F7.tt
Also, I went to MODELS in 2011 when it was in my home city, Wellington, NZ. It felt very European. I felt I could almost count the number of Americans on my hands. There were also people from Canada and South America, but they were mainly also originally from Europe too.