Tags: * against openreview * for rolling review * improve communication * EMNLP submission modalities were poorly communicated * for tracks
"(1) OpenReview is massive pile of poisonous garbage from the depths of hell. I say this as an Action
Editor, area chair, author, and ACL workshop organizer. My answer to question 9 is actually ""Keep rolling
review in some form, but do not under any circumstances use OpenReview"", but that was not an option
which is why I opposed all three choices. (2) The way EMNLP is set up this year is confusing. (It also
should not be located where it is, but that is another matter.) I had a paper ready to go for the May
ARR deadline, but I was not capable of understanding whether I should have submitted to ARR or whether
I was supposed wait for the regular EMNLP deadline. We need to have either ARR or the usual setup for
a given conference, not both. (3) On a positive note, I have found that when reviewers actually do their
reviews with ARR, the reviews themselves are noticeably better than they were ARR. (4) Tracks are really
important, and we need to reintroduce them. I keep getting assigned papers as an AC that are totally
outside my area, and I also keep seeing reviewers who say ""This is not my area, so..."". The current
system does not work. Reviewers, ACs, and authors should be able to select multiple tracks. Papers should
not be assigned outside any of the author's selected tracks."