By: O. Nierstrasz
Summary: How a program committee can discuss and accept or reject submissions to a technical conference.
You're the Program Committee (PC) Chair for a technical conference. Make the review and selection process efficient by focusing PC members' attention on whether they will champion a submission during the meeting.
You're using Identify the Champion. You're expecting 100-200 submissions, each of which should be evaluated by three to four PC members. To distribute papers to the PC, match papers to members' domain expertise.
You're using Identify the Champion. To distribute papers to PC members and maximize each paper's chance that it will find a champion, let PC members choose papers they want to review. Use Make Champions Explicit
You're using Identify the Champion. To see which papers will be championed in advance of the meeting, on the review form, ask PC members explicitly whether they intend to champion the paper.
You're using Make Champions Explicit. It's a week before the PC meeting, and most reviews are in. Order or group the papers before the meeting according to their highest and lowest scores. Do not attempt to rank papers numerically.
You're using Make Champions Explicit. It's a week before the PC meeting, and most reviews are in. Identify which papers are likely to be championed by whom, and be sure that champions are prepared for the meeting. If a potential champion is not an expert or cannot attend the meeting, take some compensating action, e.g., soliciting an extra review.
In the PC meeting, discuss the papers in groups, following Identify the Conflicts. For each paper, invite a champion to introduce the paper and say why it should be accepted. Then invite detractors to say why it should not be accepted. Finally open the general discussion, and try to reach a consensus. If there is no champion, the paper should not be discussed.
You're using Champions Speak First. Papers authored by PC members should be accepted only if there is at least one champion and no expert detractors.