Comment 076

next comment

Tags: * return to pre-ARR system * for single-conference rolling review * for discussion, author-reviewer and/or reviewer-meta/reviewer

"Many of the options being considered here (i.e. more reviewing time, re-introduce tracks, allow change 
of reviewers / AEs) are simply crying out for going back to the original reviewing system. What's the 
point of retaining the ARR / OpenReview system? As a person who's involved in more than one community 
(although ACL being one in which I have a strong involvement for 15+ years), they uncertainty of ARR 
and inability to progress with a fixed set of reviewers, has been putting me off and I have shifted 
to submitting to other non-ACL conferences more than before. Conference-specific rolling reviewing systems 
are working well (e.g. CSCW and ICWSM using PCS), but the ARR multi-conference rolling review system 
is too complex. In my opinion, CSCW / ICWSM work well with the rolling system because SPC / meta-reviewers 
need to write specific advice on how to revise a paper prior to resubmission, so the authors know what 
they need to do (and whether it's doable). With ARR, one doesn't know if reviewers are willing to be 
more positive on a revision, and what that revision should contain. Another important weakness is that 
there doesn't seem to be any reviewer discussion in ARR (or there hasn't been at least where I've been 
a reviewer). I also believe that Softconf is much more mature than OpenReview. In short: let's figure 
a better single-conference rolling reviewer system - in the meantime, please let's switch back to Softconf 
reviews."