Tips on Organizing Workshops and Symposia

Marie desJardins

mariedj@cs.umbc.edu

Abstract

This informal document summarizes some things I've learned (through sometimes painful experience, both as a participant and as an organizer) about organizing and running workshops.

Please let me know if you find this useful, or if you have any other suggestions that I could add to future versions.

Call for Papers

Be sure the call for papers is clear-that readers will know what areas you're interested in, and what types of papers are welcome.

Give authors clear directions about writing papers (e.g., include a section on relevance to the workshop topic).

Encourage people to submit and present position papers or summaries of relevant research. This broadens a workshop, and helps to give everyone some common ground on which to base discussions.

Ask people you know who are doing interesting research in the area to submit papers, or to participate in a panel or give an invited talk if they haven't submitted a paper formally. This is a good way to get people to attend your workshop!

Planning the Schedule

Try to be creative with the format and structuring, but be prepared to jump in if your creative ideas don't work out. Some possibilities:

  1. Send out a list of questions in advance. Invite anyone who wants to, to prepare a slide and talk about an issue for 5 minutes.
  2. Break out into working groups to discuss and prepare a short presentation on a specialized topic of interest.
  3. Pose a specific problem (application) at the beginning of the workshop. Have people talk about it and work on it throughout the symposium (e.g., how would different approaches be applied to solve the problem; what are the hard aspects; how well would methods discussed at the workshop scale to a large problem).
  4. Instead of (or in addition to) having authors present their own work, have a discussant read it and present a summary, then have an open discussion with the author participating.
Leave plenty of time for moderated discussion and questions after talks. The goal should be exchange of ideas, not merely presentation of results (which would be appropriate at a larger conference). Often workshops work better with fewer people giving talks, and more time for discussion, rather than inviting everyone who submits a paper to speak, and having dozens of 15-minute talks.

Panels work better if (1) there is a well-defined topic, and (2) presentations are limited in length, with the focus on general issues rather than panelists' individual research work.

Invited speakers often work well, but be sure they're interested in, and knowledgeable about, the specific topic covered by the symposium.

Be sure that presenters know well in advance what the format will be (length of presentation, time for questions, how the discussion will be organized) and what context they're presenting in (are there other similar or related papers being presented in the same session?)

Try to group papers by subtopic or issue, and have the session chair give a short overview of the papers, what important issues they address, and how they relate to the topic at hand.

Running the Workshop

Arrange for the workshop chair or another ``distinguished member of the community'' to give an overview talk at the beginning of the workshop. This helps to define a common vocabulary and shared set of issues, which is particularly valuable if you people from diverse research communities are participating in the workshop, or if the topic of the workshop doesn't have a commonly agreed definition.

Try to keep things on schedule. If you've scheduled well, there will be some slack time in the schedule for open discussion that will help if speakers run over. But try not to let this happen too much, because that discussion time is often the most valuable part of the workshop.

Make sure the program committee members, at least, are prepared to ask questions and initiate discussion

Marie desJardins (mariedj@cs.umbc.edu)