Tags: * data collection should be opt-in, not opt-out
I'm appealed the anonymity of reviewers is not mentioned in the document. Asking reviewers to share
their reviews for a corpus is akin to asking them to disclose their identity. I shouldn't be asked every
cycle whether I want to contribute my reviews to the data collection. The default should be anonymity-preserving
and the anonymity should be part of the ARR thought process. People will retaliate professionally if
they find which reviewers are rejecting their papers.