Friday, September 09, 2011

Guest Post on Conference Locations

(Samir Khuller Guest Post.)

On Conference Locations:

I recently looked at Orbitz for fares to Japan for SODA 2012 in Jan. The round trip fare from DC to Tokyo is close to $1700. Together with the train to Kyoto, we are looking at a $2000 travel cost to attend SODA for 3-4 days. Together with the registration and hotel, I am sure the cost will exceed $3000. I wonder how many people will be able to attend this conference? In times of declining budgets, we should make these decisions after careful thought since our travel budgets are only shrinking. In 2012, all conferences are likely to be expensive to attend -- SODA in Kyoto, CATS in Melbourne, STACS in Paris, Complexity in Porto, ESA in Slovenia, ISMP in Berlin, ICALP in UK, SWAT in Finland, LATIN in Peru, SIGMETRICS in UK, FST&TCS in India, not to mention all those exciting Bertinoro and Dagstuhl meetings. At least STOC is in NYC!

I am sure that the same dilemma is faced by people in the far east when we hold conferences in the US. However, I will be curious to know what the numbers look like. Are we going to reduce costs for 25 students who otherwise might not be able to attend SODA because its not in Japan (I hope this NOT the case, and the numbers look much better)? However, at the same time we might be making the cost prohibitively high for 100 students who could have attended the conference, but are not going due to the high cost. Wait, we did this once. We had FOCS in Rome! According to this blogpost it looks like only 172 people registered for FOCS 2004. Given that most likely 100 of the attendees were authors, the drop in attendance of non-authors is by a factor of 50% since FOCS most likely gets close to 230 attendees.

I am for the argument that once in a while its not bad to move a conference around to help people attend who normally might not; but we could explore other ways of helping such people attend. Once we move a conference to a place where not much local population will attend, its a problem (FOCS 1991 in Puerto Rico). SODA in Israel or Germany makes more sense to me since a large part of the algorithms community is from those places. One way to help defray the costs is to use part of the conference funds to help provide travel support for people whose cost to attend would be too high. Co-locating meetings might benefit us more (FCRC, ICALP-STOC 2001 in Crete). If we want to maximize interaction among people we should aim to have one large meeting as opposed to lots of small ones - the ISMB and ISMP conferences do this very well. More edges in a clique of 1000 nodes than 10 cliques of 100 nodes each. ALGO in Saarbrucken makes sense to me. High density of researchers, several meetings are combined into one along with ESA. Frankfurt is easy to get to from a lot of places in the world (except for the folks in Australia, NZ and Hawaii), and there is great train service from the airport to Saarbrucken.

One of the cheapest conferences I attended was the SIAM Conf. on Discrete Math at the University of Toronto campus. Very low registration cost and we could stay on campus in the dorms for $20/day. Having looked into (and having organized) conferences recently, I know that the dinner can cost $100/person, and the coffee break $20/person. Do we really need to have academic conferences at such large hotels and hard to reach places? Why not have a conference that encourages participation, as opposed to one that discourages it? Even I would not mind a conference in space, lets see if NSF would approve "foreign travel" for that one.

I am going to have to start a new US "regional" conference, that I can afford to send my students to! It will be held on a university campus and the reg fee will be $100/student; and it will be cheap to get to. At least for the years when SODA is not in the US, such a meeting might be a success.

NOTE: I have nothing against Kyoto and Rome, they are among my favorite places in world.


  1. Your 2012 list did not include FOCS. We are still soliciting bids for FOCS 2012 so there is one more opportunity to get another major conference close to home next year. (Anyone interested in hosting FOCS should send mail to me and/or David Shmoys.)

  2. But how do you explain such a large number of submissions to this SODA in Kyoto?

  3. Kayak > Orbitz. I just found a roundtrip Dulles to Tokyo ticket with 1 stop in JFK on kayak for $1548 (Delta).

  4. "Are we going to reduce costs for 25 students who otherwise might not be able to attend SODA because its not in Japan ... at the same time we might be making the cost prohibitively high for 100 students who could have attended the conference, but are not going due to the high cost."

    The nice thing about averages is that they are taken over time. Having the conference every year in Japan is probably a bad idea. Having it once every 5 years outside the US seems like a reasonable solution. Not having it ever, in over 20 years, outside north America (it was once in Vancouver) seems like a terrible idea that only people from the US (or at least just Samir Khuller) support.

  5. How many submissions for 2011 versus 2012?

  6. If you go by averages then if you shoot at a target and miss to the left and then shoot at the same target and miss to the right you can declare "Got it! but you still didn't hit the target.

  7. When was the last time ICALP was in the US?

  8. FYI, my understanding is that the ICALP steering committee recently tried to organize ICALP in the US, but couldn't get anyone to volunteer to do the local organization. If you are willing to do the local organization for ICALP in the US, I suggest you contact the ICALP steering committee

  9. ICALP? As often as STACS and a variety of other "international" conferences. At least ESA is honest.

  10. "In 2012, all conferences are likely to be expensive to attend -- SODA in Kyoto, CATS in Melbourne, STACS in Paris, Complexity in Porto, ESA in Slovenia, ISMP in Berlin, ICALP in UK, SWAT in Finland, LATIN in Peru, SIGMETRICS in UK, FST&TCS in India, not to mention all those exciting Bertinoro and Dagstuhl meetings. At least STOC is in NYC!"

    In 2012, all conferences are likely to be cheap to attend. STOC is in NYC. FOCS will likely be near a major airport in the US. As will RANDOM, APPROX, OSDI, COLT, NIPS, Snowbird, Crypto, and NYC theory day. And also all those workshops at Princeton, Stanford, Berkeley, MIT, MSRI, etc.

    More seriously, I think you are overreacting to one instance of one conference being moved far from you. People from Europe and Asia come to US for conferences. People from west coast go to Europe for conferences. Total costs in those cases are not much less than they are for you in this one instance.

    As you know better than most, fractional solutions to location problems are usually cheaper and more egalitarian than integral ones.

  11. I am in US now and I would really like to go to Japan.

  12. I guess this generated more discussion than I expected. I think some points were missed since my post was way too long.

    1.Try to pick locations close to centers where the "demand" is. For Algorithms, I can think of London, HongKong, W. Europe (Germany, Holland), Israel etc. I am not sure even New Orleans satisfies the criteria of high local demand, say compared to San Francisco or New York.
    I am not saying other locations dont satisfy this, but the argument needs to be made, thats all.
    The argument to make should be - here is the set of people who will benefit if we hold the conference at this location, not "Oh, look, what cool sightseeing there is to do". In 23 years of going to Bus. Meetings I have never once heard the argument about asking the question - which people do we benefit by moving the conference to a certain location?
    FOCS would not have been in Burlington and Puerto Rico then.

    2. Perhaps rather than just giving voice to those people who have travel money voting at Bus. meetings, we should also give a voice to those who could not attend due to already limited travel budgets. Open the voting to SIGACT members?

    3. Why not try to lower costs the way SIAM Disc. Math has done? Travel costs are only part of the picture. Using a Univ campus substantially lowers costs.

    4. Combine meetings the way ALGO has done so nicely. Giving us fewer but bigger conferences to go to, thereby increasing interaction.

    To Anonymous - I never once suggested the idea of not moving conferences around. Please read the post carefully. I even suggested locations around the globe.

  13. Japan, great! It is much better than NY or SF even for people in US,
    except Samir.

  14. COLT 2012 is in Edinburgh, actually. Not in the US.

  15. > 3. Why not try to lower costs the way
    > SIAM Disc. Math has done? Travel costs
    > are only part of the picture. Using
    > a Univ campus substantially lowers costs.

    It's how usually the conferences are done in Europe - all on the campus, and so less expensive.

    However, in the past the main problems with this approach in the US (STOC,FOCS,SODA) was to find an organizer for such an event. It's much easier for an organizer to ask for higher registration fees than to do thousands of hours of work to prepare a conference.

    ALGO in Saarbrucken had a good number of local organizers who were doing the job. They registration was lower, but the hidden cost for MPI were quite big.

  16. > 4. Combine meetings the way ALGO has done
    > so nicely. Giving us fewer but bigger
    > conferences to go to, thereby increasing
    > interaction.

    FCRC was an event of that sort. However, since it was a "commercial" the registration fees were very unfriendly; to attend (register to) 2 conferences one had to pay 1000$ registration fee, and so very limited discount.

  17. Main theory conferences still have much lower registration fees than WWW, or INFOCOM, despite many commercial sponsors. And they draw a bigger number of participants.

  18. From a Finnish perspective, location does not affect the cost of attending a conference that much. A conference in Europe costs roughly 1500 EUR to attend, while a conference in North America or East Asia costs 2000 EUR. This includes registration fees, hotels, travel, daily allowances and other expenses the university has to pay.

  19. And nobody asks the main question: Are in this "internet world" conferences not a dead-line? We now send submissions via web-platforms. We referee papers (also for conferences) via them. A great change when compared to old times of snail mail. So what is so special about conferences? Is the wish to travel, to see new countries and to meet people not the main reason why we stick of this "old fashion"? I don't speak about small workshops or the like (where you *really* meet people), I mean these "monsters", like FOCS/STOC/(long list)/ICALP/STACS/MFCS/FCT with their overloaded programs. Is a *physical* presence (costing lots of money) still necessary? I don't complain, I just wonder.

  20. Are strictly in-person conferences so "last century"? I would say yes. There are many things that make travel to conferences problematic for different people. Finances, teaching duties, family obligations, a dislike of travel, difficulty with visas, etc. I see no reason why I should not have the option of giving my presentation and taking questions via technology. Would I "miss out" on the social aspects? Only if we don't use technology to support that and only if I consider that as something on which I am "missing out" by not being there.

  21. I support the idea of making conferences cheaper to attend, and in particular to try to use university facilities if possible, but just a few notes:

    Size: the expected number of attendants is crucial. If you have at most 250-300 people, there are facilities at many universities that could be used. Beyond 300, it gets much harder to do it at US schools. (As an aside, catering costs at US schools are sometimes held too high by agreements between schools and exclusive providers.)

    Costs/size: large conferences tend to get more expensive per person. Reasons are probably fewer competitors locally (there are only one or two places in town and they tend to be high-end), and that larger conferences tend to get more professional help when in comes to organization. This seems to hold across communities.

    Volunteer approach: I think it is great if costs are held down by using volunteers. But how many hundreds of hours are you expecting them to work to save everyone $50 in registration? I also think people in the US seem to take more of a business-like attitude to this question, being more organizers than hosts. Not judging this, just observing.

    But if the community wants to have more conferences on campuses, we need to (1) consider what mix of large and small conferences we want given that the larger ones will likely not be on campus (2) make sure the conferences are outside the semester of the school hosting it (much easier to get conference and dorm rooms outside the semester, esp. for multiple tracks), and (3) most importantly, volunteer to actually do it.

  22. Sorry, this last "anonymous" (of 12:26 PM, September 10, 2011) was me. Would be happy if Bill and Lance just turn off this "anonymous option". Writing a comment means "being ready to fight for this", not less. When reading "anonymous comments" I remember about these dark times in the USSR. Anonymous could send you to jail. Just with one comment you are not a "real soviet".

  23. > Is a *physical* presence (costing lots of money) still necessary?

    A good question. All STOC talks are available online, same with FOCS 2010. And so perhaps we should try cheaper options: see the papers online, and watch the talks online.

  24. "see the papers online, and watch the talks online"

    more importantly, GIVE the talks online

    it never seems right to me when the people creating the content have to pay for the "privilege" of sharing what they've created with the community so that others can make a profit

  25. The STOC 2011 talks are online ? They appear to be behind a firewall. Is there another link that is publicly accessible?

  26. One of the reasons FOCS in Rome got such a low attendance is that since FOCS is usually in the US, Europeans do not have a habit going and attending this conference. If it were more often in Europe, it would get local loyal attendance.

  27. I may be an exception, but FOCS 2004 is the only FOCS I ever attended, and Rome was a big factor. It allowed me to meet several researchers (who do not go to my "usual" conferences), which proved to be very helpful later on.
    I'm based in Europe, and I try to minimize traveling to the US.

  28. -- SODA in Kyoto, CATS in Melbourne, STACS in Paris, Complexity in Porto, ESA in Slovenia, ISMP in Berlin, ICALP in UK, SWAT in Finland, LATIN in Peru, SIGMETRICS in UK, FST&TCS in India

    I'm surprised this thread has gone so far without anyone posting the obvious:

    Most of those conferences are being held in their normal geographical location: the Australian conference in Australia, LATIN in Latin America, the Indian conference in India, the European conferences in Europe, the Scandinavian conference in Scandinavia, the French-German conference in France...

    Even Complexity is being held in its normal place since it regularly meets every third year in Europe.

    The only conferences which are being held outside their usual continent are:


    that's it.

  29. Hear, hear, Samir!

    In the interest of reducing costs and fighting CO2 pollution, you could start a "SODA mirror site" and organize a workshop atthe same time as SODA, dedicated to reading and presenting SODA papers, but that would happen on the East Coast somewhere.

    How about that?

  30. I am going to have to start a new US "regional" conference, that I can afford to send my students to!

    I'd say it is about time. I cannot think of a single US based TCS conference being created in the last 15 years (EC, the accidental TCS conference doesn't count).

    To the contrary, US based conferences have refused to grow along with the field. They have nothing but themselves to blame if they find themselves numerically overtaken by conferences overseas.

  31. It is astonishing to see such selfishness. United States are not the center of the world, and other students also have the right to attend conferences.

  32. Let me thank STOC and FOCS for recording the videos of the talks. They are very helpful even when I attend the conference.

    I am all for cheaper conferences at academic institutions. In particular, the important conferences like FOCS and STOC should be as cheap as possible so students can attend.

    We can use students as volunteers. A few weeks of work won't kill anyone. That is what they do in some system's conferences.

    Is attending a conference in person still useful? Yes, definitely, even more so for non-senior researchers who want to built their research networks.

    Should it be possible to present a paper without physically attending the conference? I think so. It should be possible to record the presentation ahead of time and send it to organizers. They can post it on the conference website and possibly play it at the conference. That may even increase the quality of presentations by removing the show/stage factor of it.

    Maybe SIGACT, ACM, or IEEE should employ a few people to just handle the technical and organizational part of the work to reduce the time spent by the hosts on such matters. We may end up having more options when deciding the locations.

  33. How about a decent digital video-conference system?

    Talks would be fairly straightforward (one-to-many video streaming), but we could maybe use our screen, webcam, mouse and keyboard to navigate a virtual lounge during coffee breaks... We could have the conference dinner at a virtual dining table (eating our own home-made food in front of the laptop)...

    That would be neardy, hun?

  34. "It is astonishing to see such selfishness."

    I agree. Look at all of those conferences in Europe that never go to the US to give those students a chance to present there or hear those speakers.

  35. Conferences held during the academic year (e.g. SODA, FOCS) can't be run on campuses. (A few northeast universities are closed in January but they are really closed then and not able to host SODA.) The difference between holding something on a campus versus a hotel is at most $100 and probably less unless people are actually staying in dorm rooms.

    If you want a conference that is convenient/cheap for your geographic area then you need to get someone to VOLUNTEER to host it in your area. Compare SODA and FOCS/STOC registration costs - FOCS/STOC are run by volunteers, are a lot cheaper AND provide lunches for the full conference. The main reason that conferences with very large attendance are costlier is that less can be done by volunteers.

    Claire and one or two others argue that conferences are out-dated because of the travel impact and cost. The number that we have IS excessive. However, it seems that there is something different about the influences and mixing of ideas one gets from sitting in a talk on something one didn't know anything about before and seeing some connection - or simply hearing that something is important or worthwhile from the chatter in the hall. Here is a simple test - how many recorded talks do you watch from a conference you are not attending/

  36. @Paul: I am not advocating for "conferences at a laptop". Meeting people, exchanging ideas "face-to-face" is very important, no question. But my experience was that I can only get all this "face-to-face" in relatively *small* workshops, like Dagstuhl-Seminar. (I do not count a 3-5 minutes talk with someone somewhere in coffee break.) 3-4 days long conferences with about 300 and sometimes much more participants is not the same. Costs = ca. $2000 for ones budget per conference, and tons of CO_2 for the environment. Big advantages = I don't know. One more aspect: how many "surpises" does one experience in conferences? I mean, results that one has never heard before, that one has not already received from colleagues or has not already fetched from their home pages. At least for me, the results of *all* talks of interest were already known, before the conference (via direct contacts); this is why I am no more interested in attending "monster conferences".

    What I am trying to say is that the current format of conferences for *senior* scientists is a dead-line (the fraction of merits/costs is too small). The situation with *junior* scientists is somewhat different. For them, conferences have another, more pragmatic aspect: meeting people helps for their own career. So, for now, they need conferences.
    And this is the main reason why I am *for* conferences, but for *cheaper* ones.

    Finally, I fully agree with Lance and others that "CS should grow up". Conference publications should be counted as *preprints*, not as real publications (as this is in all remaining fields). Many of the problems we discuss now would then be automatically resolved ...

  37. Just in case someone has missed it:

  38. Dr. Anonymous has too many opinions for me to respond effectively to all
    the comments.

    Since I have been accused of "selfishness" - let me point out a few things.
    I even suggested that the conference use (my) registration dollars to help
    defray the costs of people for whom the cost to attend is too high. I am
    simply trying to speak up on behalf of the students for whom conferences
    are too expensive to attend.
    The question simply is "why not hold a conference close to the demand?"
    be it London, Israel, Germany or HongKong. I am not saying the demand is
    not there in Kyoto, but this argument needs to be made.
    I am not sure if there is significant local demand in New Orleans either, a
    favorite SODA venue. Moreover, lack of good public transportation increases
    the costs and people are forced to stay at the main hotel. In HongKong,
    $3 gets you by an express bus from the airport to most places in the city and
    the cheap metro means you can stay in cheaper hotels and hop on the metro
    and be at the conference in 15 mins.

    I would like to visit some game parks as well in Africa, why not have a SODA
    there? The main point being that conference locations should not be decided
    based on the vacation spots its attendees would like to go to; but to make
    it easier for students to attend -- WHEREVER there is strong demand.
    If this were paid attention to, we would not have had a FOCS in Burlington
    or in Puerto Rico - both of which suffered from low attendance (and for
    some reason Rome as well).

    Organizing a conference is a lot of work - SoCG 2008 was right here in
    College Park with very low registration cost thanks to Dave Mount
    and many Ph.D. students in our department (including several of my
    students). STOC 2009 (Bethesda) was again organized by our department -
    Aravind Srinivasan, Jon Katz and Dave Mount with students helping them.
    However, this conference was not on campus and cost quite a bit more.
    Last year I organized an NSF sponsored Broader Impact conference with 125
    people. It is a big headache to organize something both at a University
    campus as well as at a hotel. However the costs are substantially different.Algorithms people in Europe have several meetings to go to - ESA, ICALP,
    STACS, FUN, CIAC, SWAT to name a few. If SODA is an international conference
    and is expected to move around all over the world, then this points to the
    need for a regional US conference, thats all. If SODA is the regional
    meeting (which it appeared to be initially), then everyone in the world is
    welcome to come here to attend, just like we attend ISAAC, ESA, ICALP,
    LATIN, CATS, SWAT etc. In any case a conf should be one that is cheap to
    attend (on a campus), without anyone paying $100 for a meal and $20 for
    coffee. Without this large cost, the reg. fee would be low and the entire
    conference can be attended by more students. I have sent several undergrads
    to SODA in the past, and the cheaper the conference is, the easier this is
    to accomplish.

    Interaction among attendees is the main benefit of an actual meeting
    rather than video lectures. This is not to say that we should
    not tape the lectures and distribute them widely, just pointing out what
    we miss when we all sit at home and watch video lectures.