Handbook chair duties

(contributed by Matt Post)

• Time frame: In general, if you can manage two months between the conference start and the main conference and workshop book deadlines, you should find yourself in good shape. So ask for 8 weeks and be happy when you get 6 or 7.

• What follows is a small guide written by Matt Post. The handbook AND publication chairs are encouraged to read as soon as possible: https://github.com/naacl-org/naacl-handbook/ This has been forked multiple times (for subsequent NAACLs, ACLs, and EMNLPs), and there may be updated information elsewhere.

• Build the handbook as early as you can, and then start updating it. The hardest part, by far, is the main conference schedule. Do your best to get as much of the rest of the stuff out of the way ahead of time. This includes various and sundry tasks such as:
  • Get a designer to produce a cover for you. 8.75" x 5.75", which includes a 1/8" bleed. This is usually similar to the website design. Don't do it yourself.
  • Get bios, pictures, and abstracts from invited speakers
  • Create pages for special events (banquet, special event, business meeting, etc). You should have a nice little writeup that calls attention to all of these events.
  • Get someone to write the local guide. These are best when they're personalized by people who know the area. For a great example, see the ACL 2014 handbook (http://acl2014.org/handbook-acl2014.pdf).
  • Get ads from all sponsors, sized appropriate to their classification (gold, silver, etc). These go in the end of the handbook (from Priscilla)
  • Get hotel maps for inside front cover (from Priscilla)
  • Pester the GC and PCs for their introductory letters
  • Think if there is anything cool and new you'd like to add

• The main thing you care about is the printer deadline (usually OmniPress), which Priscilla here has pegged at roughly T–4 weeks. Work that out with your OmniPress contact. Schedule the week before this deadline to dedicate mostly to handbook assembly. Gather your patience for dealing with inexperience workshop chairs.

• Send reminders to workshop chairs about the book deadline, which should be roughly T–5 or T–6 (if you can manage it). Workshops are a bit easier to format but see the notes in the linked guide above about getting workshop chairs to use the proper formatting in the order file. In the end, it doesn't quite matter, because workshops aren't printed out in that much detail.

• The hardest thing is the main conference. You have to incorporate TACL papers. You have to get the schedule in a properly-formatted "order" file from the PCs. This has to be finally, finally, finalized at T–5 (this is what fills the week before the T–4 deadline). This means setting a deadline well before that. Build a relationship with the PCs early, and don't get intimidated about sending them lots of increasingly urgent reminder emails just because they're more famous than you. I recommend picking up the phone at some point in the process, preferably early, just to help people get on the same page about what's needed.

• Make sure to work closely with the publication chairs. They are putting together the proceedings, which contain the schedule, but only print it flat, similar to the way the handbooks print workshops schedules. YOU (the handbook chair) need a properly formatted, machine-readable "order" file, because you have to pull out times and paper descriptions in order to build daily and session overviews. This is a real problem because the pub chairs have the power here but it is the handbook chair who needs the job done in a certain way. This is why I have argued that the pub chairs should also do the handbook, but this doesn't always happen.

• Author index:
  • Gets everyone's name correct and consistent across the whole conference (tutorials, introductory material, credits, main conference, workshops)
  • Includes index entries for everyone involved in the conference, including things like session chairs and so on. Much of this has been automated if you use the proper tags in the "order" file (which could be better documented)
  • This usually involves compiling the proceedings and then sending them to Drago, who runs them against the Anthology Network and finds corrections. This has been made easier in the past few years by the START logins which let people choose their name formatting and make it consistent, but people don't always do this. Also, the workshop and pub chairs have a duty to go through and verify these consistencies themselves, in part because it can be a lot of work, they don't always do it.