Tags: * in favor of infrastructure, but what exactly is part of it? * good reviewers do not get invited * improve matching
"In general, an integrated review pool needs to be clear to everyone *what is getting integrated* in
the first place. If the idea is for this to be an infrastructure for *ACL and EMNLP, that is fine, but
it needs to be very explicit to align people's expectations. Conferences do NOT have the same themes
or expectations, and it is unrealistic to ask reviewers to write a generic review as the notion of ""generic""
does NOT exist. Also, it is a widely known issue that people who are perfectly qualified to review,
and are KNOWN to be good reviewers in the past, do not even get invited. This needs to get resolved.
Poor reviewer/meta-reviewer assignments and late reviews are by far the biggest problems I have observed."