2009Q3 Reports: EACL 2009

      General Chair's Report for EACL 2009
               18th March 2009

My job as general chair was made incredibly easy because I had a totally spectacular team at every level of organisation. The local team especially have been fantastic. Their attention to detail has stopped me from overlooking crucial issues time and again.

The conference was the largest EACL ever, with over 500 people registered.

What follows are just a few thoughts from which I think we can learn lessons, and some recommendations for the future.

This is the first time we have had a general chair for an EACL that is not held jointly with ACL. This additional role is now necessary so as to ensure continuity in organising the triennial conference while the EACL chair changes on a biennial basis. I was EACL chair at the time when I was appointed as general chair, and this violates ACL policy about appointing general chairs. While there was a rationale for violating the policy this time in order to get the new structure in place, the EACL board should consider appointing an external general chair in future. There is no real reason why the people who choose the conference site need to include the general chair. Rather, an extensive knowledge of ACL policies regarding conferences and knowledge of the local organising team is key for the role of general chair (and so, in fact, there can be advantages to choosing the general chair after the conference site is chosen).

The PC chairs were very receptive to the EACL board's concerns about diversifying the kinds of papers that get submitted and accepted. We changed the review form to encourage this, and I think it worked well. As the PC chairs' report will show, we had the largest number of submissions ever, and the largest main programme. I think that the EACL board, and the ACL exec, need to continue thinking of creative ways to diversify papers at our conferences.

The poster session was a resounding success. My own feeling is that it would be useful to experiment with it further, particularly on increasing the propertion of papers that are presented this way. However, very unscientifically, we had a straw poll at the general meeting about the proportion of papers to be presented as papers, and that straw poll showed people thought the proportion was about right. Also, about half the people present at the general meeting thought the 10 minute talks accompanying posters was the right time; the other half felt 5 minutes would be better. All the people who expressed an opinion thought the short talks were better than having no talk at all. Overall, as a field, we must work hard to remove the stigma that normally goes with presenting one's paper as a poster. To that end, it is vitally important to give the poster session prominence in the programme, and it is also crucial that one should be unable to tell from the proceedings which papers were presented as talks and which as posters. the poster session. EACL 2009 was the first conference to ensure that one cannot tell from the proceedings the mode of delivery of the paper.

There was some discussion early on about EACL's policies on PC chairs and area chairs submitting papers. We articulated a policy and since then this has been made into official EACL policy. I would recommend that the EACL chair take forward to the ACL exec a proposal that this be made into ACL policy.

Joakim Nivre's expertise in both using START and in electronic publication was utterly invaluable. Before the publications chairs were even chosen, he was advising that all workshops should be restricted to using the START system. This was really good advice; the job of electronic publication would have been much harder if the workshops had not used START. I think we should make it policy that for all future EACLs, all papers are monitored and updated using the same software system.

The schedule for producing final papers for the main programme and workshops also went slightly awry because when I drew up the initial schedule for all the conference tasks, I underestimated the time required to produce camera ready copy from the pdf papers. Again, Joakim Nivre's expertise with electronic publication---he had prior experience as an ACL publications chair---enabled him to notice this error in good time to fix it. We brought forward the date by which final papers were due without any noticeable inconvenience to the authors. So the error was fixed before it caused problems. But this was a fluke: without Joakim noticing the mistake we would have had serious problems with delivering the proceedings on time. The lesson to learn from this is that EACL needs to extensively review the recommended template schedule for running an EACL conference, and make sure that the time for each task matches the reality.

Throughout the last 12 months, money has been a worry. There are two main sources for this, and we can learn lessons from both of them. The first is the extremely low level of non-local sponsorship. Non-local sponsorship was coordinated among all ACLs for the first time. I think that the work done by the Sponsorship working group, that was appointed by the ACL exec committee, to rationalise the levels and prices of sponsorship has been very helpful. While the seeking of sponsorship for EACL wasn't successful this year, I don't think it was because of this coordination effort. We should continue to coordinate sponsorship activities in future. However, whether sponsorship is coordinated or not, finding good people who are willing to take on this role is incredibly hard. It is a horrid job, with no real personal benefit (unlike PC, workshop or tutorials chair, which is of benefit to one's CV as it shows your international standing in the field). I think that the EACL board, and the ACL exec, needs to think about how one could provide incentives to be a sponsorship chair. In the case of this year's EACL, I feel quite sure that while many letters went out, there was very little follow through with phone calls etc. More effort was needed in these trying times.

The second source of money trouble was that very early on, we committed to an expensive conference venue. The budget for the venue, not including catering, is half our total expenditure. However, when we made the decision to move from a cheaper venue to this one, the estimated expenditure was much less than this. This was because those figures, on which we based our decision to move, involved room bookings that in the end had to be changed to more expensive room bookings. I discovered this only after a site visit: the change to the room bookings was necessary so as to ensure that during coffee breaks, lunches and the opening reception people had access to natural light. An earlier site visit would have spotted this problem much earlier on, before we were contractually committed to the conference venue. It would have given us more leverage during financial negotiations with the conference venue, and may have enabled us to acquire what we needed at a cheaper price. So in future, I feel that site visits should happen around 18 months in advance of the conference itself, rather than 6 months in advance as it is now, and if at all possible in advance of any contracts being signed with any particular conference venue.

In the end, the conference made a tiny profit, of about 2000 euros. I was relieved it wasn't a loss! And without that massive local sponsorship, far outstripping their targets (they raised 45000 euros, compared with their 20000 target), we would have incurred a huge loss.

I think that the coordinated call for workshops worked very well, and we should continue to do that in future. However, one of the workshop chairs among all the conferences needs to be appointed as the overall leader of the group, so that there is one point of contact for all workshop proposers and for the general chairs of the conference.

While the tutorial chairs did a brilliant job in acquiring tutorial proposals, from which they formed what I think is one of the strongest tutorial programmes that we've had in years, attendance at tutorials is dramatically down. The EACL board should think very carefully about what they want to do with tutorials at future conferences.

The local organising team were totally fantastic. They solved every problem that was thrown at them---some of them of my own making---in a timely and professional manner. They also had a significant creative input into the conference, and always communicated with me efficiently and effectively. The sheer time and effort that the local team have put into this conference is staggering. EACL owes them a massive debt.

One thing that may have gone unnoticed is the support not just from the actual local team themselves but also from their academic institutions. Of course, these institutions always bear the brunt of hidden costs when their members organise a conference of this size. But in this case, they have exceeded all expectations: they have not only borne more hidden costs than most institutions do (for instance, most of the web and cover design was done in house by a scientist who is particularly gifted at this kind of thing, saving us a fortune on the publicity and printing budget), but also they gave cash directly to the conference in the form of travel expenses for tutors and invited speakers. The EACL executive dinner is an important event in this respect: given that the directors of these institutes don't attend the conference itself, it is the one opportunity we have to acknowledge and recognise everything that they have done for us.

Finally, I think that EACL could use Priscilla Rassumussen's expertise more effectively. She knows everything there is to know about organising a conference, and she was a great source of advice to me. If the general chair is not appointed 18 months prior to the conference, then Priscilla could do the site visit, for instance.

Alex Lascarides. General Chair of EACL 2009.