I wrote a column for Internet Computing about reviewing the reviewers. A PC member (or periodical reviewer) who agrees to review one or more papers and then either does not perform the review(s) or does them badly should be bypassed in the future, but it is hard to identify the people who do this. Pointing the finger too publicly can have repercussions. But certainly if someone leaves multiple conferences in the lurch, they should never be asked by anyone ever again. We need some sort of neutral agent to collect these reports but not make them publicly available without sufficient evidence.

-- FredDouglis - 01 Apr 2008