Small steps for better TCS conferences?

Question

Often, when we take part in TCS conferences, we notice some little details that we wish the conference organisers would have taken care of. And when we are organising conferences, we have already forgotten it.

Hence the question: Which small steps we could easily take to improve TCS conferences?

Hopefully, this question could become a resource that we could double-check whenever we are organising conferences, to make sure that we do not repeat the same mistakes again and again...

---

I am here interested in relatively small and inexpensive details – something that conference organisers could have easily done if only they had thought about it in time. For example, it might be a useful piece of information that could be put on the conference web page well in advance; a five-dollar gadget that may save the day; something to consider when choosing the restaurant for the banquet; the best timing of the coffee breaks; or your ideal design of the conference badges.

We can cover here all aspects of conference arrangements (including paper submissions, program committees, reviews, local arrangements, etc.).

---

This is a community wiki question. Please post one idea per answer, and please vote other answers up or down depending on how important they are in your opinion.

Answer 1

1. Keep prices down: you can get lovely hors d'oeuvres at a $750-per-seat hotel-hosted conference, it is true, but it tends to detract from the intellectual atmosphere you can get with a $100-per-seat university-hosted conference. Also, you don't just get the people attending who deliver talks. Allow students at the university to attend for free;
2. Space parallel sessions, so that people can really move between them;
3. But also encourage interaction between speakers and attendees at parallel sessions beforehand, so that parallel sessions are also autonomous communities. This encourages interaction, and the sense that parallel sessions are moving the discussion in their subfield forward. Some conferences (e.g., the German linguistics conference, DGfS) have taken this to the conclusion of having all parallel sessions be independent workshops.

Answer 2

A limited number of tables/desks in a quiet room

Problem: Not everybody will attend every session, nor will everybody be staying at the hotel to/in which the conference is nearest/held.

Solution: A few tables, desks, chairs in a quiet room for undisturbed reading, writing, typing, browsing, possibly with power outlets.

Answer 3

Give away USB sticks as proceedings

Yes, the papers are online, but conferences rarely have working wireless in the conference rooms or nearby public areas. CD-ROMs are the current electronic medium of hoice, but a large fraction of people have tablets and netbooks without CD drives. USB drives cost less than dirt, and they work with almost any computer whose name doesn't match the regexp "iP*".

If you don't want to pre-fill each USB stick with the electronic proceedings, let people fill their own sticks with the electronic proceedings, either from a PC (or three) at the registration desk, or from a colleague who already dumped their stick onto some other device.

Answer 4

On the conference web site, post simple but detailed instructions for getting from the airport to the conference hotel by public transportation.

Don't give dozens of options – just one or two routes that are straightforward and easy to follow. Tell what kind of tickets to buy, from where, how, and how much it should cost. Make sure that the information that you post is actually up-to-date and valid, especially if the conference takes place in summer.

(It might be a good idea to have a look at a Lonely Planet guide book to get a good idea of what information is useful for someone who is visiting the town for the first time.)

Answer 5

Don't schedule talks about similar topics at the same time.

Self-explanatory. There's no excuse for it.

Answer 6

Make sure that WLAN actually works, even if all conference participants log on to it simultaneously during a coffee break.

Further reading:

• The “WiFi At Conferences” Problem, by Joel Spolsky.
• Why is Internet access and Wi-Fi always so terrible at large tech conferences? @ Server Fault.

Answer 7

Abstract booklets.

Problem: proceedings are now mostly online (a good thing) and I don't feel like uncorking my laptop to see the topic of a paper.

Solution: Supply abstract booklets (some conferences are already doing this). It's a small booklet that lists abstracts in order of presentation and is an easy reference to decide what a talk is about.

Note: this is an interim solution till we all carry around IPads :), but even then having all abstracts summarized in one place is very convenient.

Answer 8

Calendar for conference activities

Problem: It's hard to keep track of multiple sessions, and often in a large conference I don't want to attend all the talks, or would like to maintain a special list of talks I'm going to.

Solution: The conference should at a minimum provide a calendar file that can be slotted into google calendar or whatever calendaring app you have on a phone/laptop. The ICS format is pretty standard.

Note: Conferences can go much further - at KDD you logged into a social networking site associated with the conference and could export a customized calendar based on what talks you're interested in. But that requires more infrastructure and money, and is not within the scope of this question.

Answer 9

Get rid of 2-column format, and generally streamline paper preparation

Right now the typical TCS paper has 4 different versions:

1) A preliminary full version - this is the version with full proofs you work on before the conference submission, and the one many people post online around the same time as submission.

2) A submission version - this is a version, hastily prepared a couple of days before the deadline, where we convert a nice 30 page paper into a less readable 10 page paper + 20 page appendix by copy/pasting proofs into appendices.

3) Camera-ready version - this is the version in the two column format that actually gets published in the conference. This is obtained by incorporating reviewer feedback and insights we obtained in the meantime into the submission or preliminary full version, and then working to make it compatible with the ugly 2-column format template provided by the conference. (I think there's more or less universal agreement that the 2-column format is ugly, and the only reason it exists is as a remnant from the not so long ago days of printed proceedings.)

4) Journal version - this is a full version that also contains the above feedback and insights, as well as the result of another pass that we submit to a journal. For technical reasons, it's often hard to maintain the same tex file for versions 3 and 4, which means that one needs to duplicate the work in creating these two versions.

Given this, perhaps it's not surprising that many papers don't make it to stage 4, and the most mature version that remains is this ugly 2-column format.

I'd suggest the following changes:

1) As Emanuelle said, submit a full preliminary paper (i.e., version 1) to the conference, with the understanding that the PC is not obligated to read beyond the first 10 pages, and that they generally can use their judgment to skip technical details of the proofs.

This makes it easier for the authors (one less version) and also for the PC, since if they actually want to see the proofs, they don't have to hunt for them in the appendices, and they can always simulate the current situation by just printing the first 10 pages of the submission and ignoring the rest. (Russell Impagliazzo once said that a submission should have no page limit, but just an instruction that for x=1,2,4,8,... reading the first x pages should make the reviewer want to read the next x ones.)

2) For the camera-ready version, use a 1-column format with a minimally intrusive latex template, that will make it easier to keep working with the same file.

In fact, what I would suggest (though it's a bigger change) is to make the camera-ready version a true "extended abstract" by restricting it to 10 pages of single column. Given that your paper is already accepted, and that a version with the full details should be available online, you don't have to waste space in these 10 pages on explaining more details or enumerating all arguments why your paper is great, and you can focus on trying to explain in the simplest possible way the main ideas behind your work. I think a paper in this format will actually be a valuable resource and will justify having one more version on top of the full/journal version. Also, I think the fact that the conference version will be necessarily be without proof details will encourage people to submit to journals.

Answer 10

Recommend a few good hotels to conference participants

I've been to conferences where the recommended list of hotels had 10+ hotels (or none, with the message "this is a large city with many hotels, feel free to pick any"). While it's nice to have so much choice, first it means that someone who doesn't know the city will have to pick which one is optimal for them. Second, this greatly reduces the chance of being in the same hotel as a lot of other conference attendees.

Instead, conference organizers, who are locals and can evaluate hotel optimality much better than attendees, can recommend something like 3 or 4 hotels at most, based on their price. There should be a cheap option for students and people on a low budget, and another one for those without funding constraints. That way a lot of attendees will end up at the same hotel and have the chance to interact during breakfast, commute to and from the conference together, etc.

Sometimes this suggestion is not feasible when the hotels in the area are small and will get full really fast. I mean to direct this suggestion to conferences where it is feasible to make such recommendations.

Answer 11

Keep prices down, but try to include the conference banquet in the student registration fees.

It seems to be all too common that students miss the banquet for a technical reason: the registration fee does not include the banquet. Their instructors and/or universities would be happy to pay the registration fee – whatever it is, and no matter what it includes – but for some reason they refuse to cover any "extra" costs.

A conference banquet could be a great opportunity for students to get to know people, and I feel sorry for the students whenever they miss this opportunity. The extra cost (on top of flights, hotels, registration fees, etc.) would be usually insignificant, so it is mostly just a matter of bureaucracy.

To keep the registration fees reasonable, it is possible to save money somewhere else – things like conference bags are much less important. You can also try to find a company to sponsor the dinner for the students. And the banquet does not need to be that fancy – great food and an amazing location would be a nice extra, but in the end the social aspect is what matters most.

Answer 12

Post a printed program near the door of the session room(s).

Sounds like a basic thing, but for some reason most conferences do not do this. It costs nothing, but would be very helpful for answering quick questions such as "Is this the session where I want to go?", "What time the coffee break ends?" without looking at the abstract booklet.

Answer 13

Conference submission = arXiv ID.

Nowadays, a conference submission is usually a PDF file. Instead of asking people to submit PDF files, we could easily ask them to just enter the arXiv identifier of the work that they want to submit.

[Of course the arXiv paper must be formatted like a conference submission: approx. 10 pages of text + all missing details in an appendix.]

This would be relatively straightforward to implement, and it would have numerous benefits:

• If a paper is accepted, this way it is guaranteed that a version with full proofs is permanently available somewhere. We would never again have the typical situation that a conference paper claims a result and refers to a future journal version that will never appear.

• All results would be made public much earlier. We could have a faster advancement of TCS, as others could more quickly build on top of earlier work, and less risk of reinventing the wheel many times in parallel. ArXiv gives reliable timestamps, so there is no question of anyone stealing someone else's work.

• Reviewing would more interesting, and there would be fewer ethical issues. Now you would be reviewing something that is already made public. You can perfectly well already start to think how to use the results in your own work, how to improve them, etc.

• When the list of accepted papers is published, it would be trivial to add arXiv links to all accepted papers. This way people could already have a look at the papers that look interesting well in advance before the conference.

Obviously there are some drawbacks, too. Posting your paper on arXiv takes some time. You would actually have to first submit to arXiv, then sleep on it, and finish your conference submission on the following day (if it still looks like a good idea). But I don't think this would necessarily be a bad thing...

---

(Of course, now it would also be obvious that there would be very little need for any printed proceedings or commercial publishers. After all, the list of accepted papers with arXiv links is, in essence, everything that we need. If someone really wants to get a fancy proceedings book, a local university press could handle it at a low cost. But this is getting off-topic – getting rid of publishers is not a small change that people could implement easily.)

Answer 14

Don't have multiple web sites for the conference.

If there are many possible URLs, make sure that all of them redirect to the same place. Don't try to maintain multiple sites by manually copying information from one place to another.

This may sound ridiculous – why would anyone set up multiple web sites with different content – but for some reason it seems to be happening all the time. There is the "official" web site with a cool domain name and a fancy layout, but it is too difficult to update it; then another web site is set up, and soon nobody knows which site to check for the latest information.

People don't need fancy web sites; they want to find information.

---

(If you haven't seen this before, here is a recent example: 1, 2.)

Answer 15

Have student reps

To address Jukka's supplementary request made in a comment to the question:

Enlist two (or more) people -- one an energetic young researcher (student or post-doc) and one an engaging and extremely experienced academic (professor) -- to co-ordinate activities for early-stage researchers. The former would be responsible for planning drinks and other social events, while the latter can provide careers and research advice, perhaps in a special-purpose session of the conference or possibly at said drinks. (An optional third person could be an industrial contact.)

Answer 16

Do not require "One Full Registration Per Paper." This forbids students from paying student registration if they have a student-authored paper. It seriously offends many, especially since this policy is usually announced after paper acceptance.

Conferences that have done this recently include ISAAC, IWOCA, and COCOA.

I imagine the reason for doing this, particularly in a conference not backed by a professional society, is to guarantee a certain revenue amount. But, it is really the organizer's responsibility to figure out reasonable projections of the student/faculty mix and set registration accordingly.

Answer 17

I summarised a few thoughts after arranging ALGO 2009, see Post ALGO post: etiquettes and etiquette. If I have to highlight only one idea (not already mentioned on this wonderful thread): visible, legible name tags, together with repeated, strong signals from the organisers to actually wear them. In particular, I think the organisers can ask the bigwigs in the programme committee, the invited speakers, and the steering committee to do so.

Answer 18

Record the talks/lectures and make the videos available online.

Answer 19

Do not restrict the length of submissions. Instead, ask authors to make their points within X pages, and inform them of the length of camera-ready versions. While some recent PC have courageously done so, this does not yet seem standard.

Answer 20

Publish Information about Grants, and Try to Allocate them at Registration

It would be nice to know what grants are available to students, new postdocs, etc., before registering for the conference and purchasing flights. (There's nothing more frustrating than finding out that you've been offered a grant to pay for attendance and accommodation for an entire week after you've planned to attend for a single day and booked flights and hotels accordingly.)

Similarly, it would be nice to get a grant at registration time, if possible, rather than to be given a cheque or cash upon arriving to the conference, or worse, waiting weeks after the conference to receive a cheque. Anti-bonus points for a conference in another country with a different currency.

(I've been to one conference where they didn't process registration payments until the day of the conference, and handed back cheques to those who received grants, so there are alternative ways of handling this.)

Answer 21

Reduce the number of sequential talks. Three half-hour highly technical talks following immediately after one another is hopeless. Nobody can concentrate for that long without a break. Ideally there should be a break after every talk.

Answer 22

Use specialised software to manage submission of papers, assignment of referees, receipt of referee reports, gathering PC member assessment of reviews, and communication of outcomes.

Conference management software on the Edutech wiki lists some options.

Answer 23

During talks, display electronically the time left to the speaker (see videos of TED talks for an example). I used a simple java application in full screen to do so at the last conference I chaired: it would have felt stupid to do it "by hand", keeping an eye on the clock and using papers with "10mns left", "5mns left", "1mn left", "FINISH NOW" written on it.

Bonus: display the overspent time in color flashing after the timer is done.

Answer 24

Put the deadline on Friday. This will make many significant others happier.

Answer 25

If you use conference bags, make them recyclable or reusable:

If you need a bag to give registrants, then you are giving out too much stuff---everyone just throws it away and often hotels do not even have paper recycling! Still, if you do use a cloth bag then choose a useful design, for example a grocery shopping bag. If you use a plastic bag, then ask the hotel for plastic recycling bins.

Answer 26

Make conference T-shirts using a large collection of figures from the papers. (ACM SCG has done this twice in its 25 years; know any other conferences that have?)

Answer 27

Include among important dates the date for special issue invitations.

Otherwise, how long are authors supposed to wait for?

Answer 28

Have sensible deadlines:

1. Submission deadlines: I understand that there are different reasons for different submission deadlines, such as flaky servers that need personal attention. Whatever the deadline is, though, post it clearly and convert it to a familiar time zone for most of your submitters (such as Eastern time, or ET, in the US). A good philosophy for conference websites is that if you can save n people each 1 search query, then please do!

2. Registration deadlines: For early registration deadlines , the only rational deadline is the end of the day, in any time zone. That is what is announced on blogs and gets put in calendars. There is no reason to make it 3pm ET (ahem).

Answer 29

Encourage CouchSurfing as an accommodation

I am not saying to bind people to CouchSurf, however imagine that with the hotel recommendation you add a page where local researchers (or student as I imagine this would be mainly used by students) could post a message saying "I can host so many people, mail me".

Only people ready to do it would use it. For example, for students going to the conference it may be a good way to:

• Lower the entrance fees (you do not pay for accommodation) (for the couchsurfer)
• Not a random CouchSurfing, meet people from the field, have a better interaction (for both couchsurfer and host)
• Visit the places to see in the city with someone that knows it, enjoy it to the fullest.

I imagine that this can be a lot of work to do from scratch, however it may be possible to have an organization specialized in CS to do this for the conference organizer since it does some advertising for them with not a lot of work. Win-win.

I know that this will not be used by every one, but I believe that those using it will really benefit from it (personnal experience, I always had a great time CS).

Answer 30

(this is not specific to TCS conferences, but would work for better conferences in general)

A nice idea I saw in a mathematical conference for young researchers is asking every participant to write a short "research statement" - in case of junior participants who don't yet have much results, a description of interests would be OK. Then, some time before the conference, the statements are published on the webpage. A working example of this: http://bcc.impan.pl/12Young/uploads/statements.pdf (conference in geometric group theory)

I think this would be especially helpful for junior participants suffering from the "I don't know anyone here" problem, but also the other way, since even students aren't anonymous mass anymore. Of course, this is rather feasible for smaller conferences/workshops, not 400-people events, but still I think it's worth implementing.