« Back to Summary
Do you think the ICML2010 reviews were different in quality from the reviews at previous ICMLs?
#Response DateAdditional comments:
1Jun 10, 2010 8:27 PMActually, I had a very "good" mix of qualities. As a rule of thumb, the better I knew the reviewer, the better the review.
2Jun 10, 2010 10:51 PMI've often found ICML2010 reviewers incoherent----making claims that just don't compute. That was true this year also.
3Jun 11, 2010 2:30 AMThe main issue seems to be the lack of reviewers with expertise.
4Jun 11, 2010 3:03 AMI thought there was a lot of poor matching -- the process is too complicated, and it depends on area chairs guessing good keywords, and having a pool of reviewers that is "right" for the set of papers they actually get -- which might or might not happen (and often didn't, especially for ACs with broad interests). I also thought that the Phase 2 reviews were no better than the Phase 1 reviews, and the shortened timeline caused by the two-phase reviewing just made it less likely that people would have the time and attention to really do a good job.
5Jun 11, 2010 1:03 PMCan't tell, but quality varied a lot!
6Jun 11, 2010 2:13 PMI can't compare very well, this was my first time as an area chair for ICML.
7Jun 12, 2010 1:08 AMI actually thought that most of the reviews on the papers I ACd were better than what I was used to receiving. But the ones I received on my paper were not that great (the paper got in so this isn't totally about acceptance - I just felt they weren't really experts in the appropriate area)