Comment Summary

Do you think the ICML2010 reviews were different in quality from those at related conferences?
#	Response Date	What conference are you comparing to and what is the subject area?
1	Jun 10, 2010 7:06 PM	this year's ecml/pkdd
2	Jun 10, 2010 7:10 PM	NIPS
3	Jun 10, 2010 7:15 PM	Reviews everywhere have a lot of variance. Hard to compare to an 'average'.
4	Jun 10, 2010 7:21 PM	aaai
5	Jun 10, 2010 7:26 PM	SIGIR
6	Jun 10, 2010 8:23 PM	NIPS
7	Jun 10, 2010 8:24 PM	ICDM
8	Jun 10, 2010 8:25 PM	my comparison is with graphics conferences, e.g., eurographics, siggraph
9	Jun 10, 2010 8:34 PM	AISTATS-2010.
10	Jun 10, 2010 9:03 PM	NIPS, AISTATS. I had one outstanding ICML reviewer (negative review), one good one (positive review) and one really poor one (negative and nonsensical). I submitted a paper that was between areas (Bayesian nonparametrics and stochastic optimization) and had difficulty getting capable reviewers. There needs to be a way to get papers like this split between area chairs and possibly have a feedback/rating system for reviewers (those with low marks would not get asked to review again in a certain area).
11	Jun 10, 2010 9:05 PM	ecml pkdd 2010
12	Jun 10, 2010 9:06 PM	ECML, GECCO, AAMAS, AAAI Reinforcement learning
13	Jun 10, 2010 9:26 PM	AAAI, IJCAI
14	Jun 10, 2010 9:56 PM	UAI 2010. Subject Area: Gaussian Processes
15	Jun 10, 2010 10:48 PM	NIPS, Siggraph.
16	Jun 10, 2010 10:51 PM	I find NIPS marginally worse as far as craziness. COLT reviewers seem to be stronger, but their viewpoint often doesn't make sense w.r.t. what is significant work.
17	Jun 10, 2010 11:59 PM	E.g., COLT 2010, reinforcement learning
18	Jun 11, 2010 3:03 AM	AAAI and SDM machine learning papers
19	Jun 11, 2010 6:21 AM	CVPR, ICCV, NIPS -- vision & graphical models.
20	Jun 11, 2010 8:09 AM	CVPR, ICCV, NIPS. Computer Vision and Machine Learning
21	Jun 11, 2010 8:20 AM	first submission in this type of conference...
22	Jun 11, 2010 8:42 AM	ECML - Machine learning
23	Jun 11, 2010 11:32 AM	ACM KDD, ECML PKDD
24	Jun 11, 2010 1:03 PM	Again, due to variance I don't dare to generalize...
25	Jun 11, 2010 1:19 PM	Same remark.
26	Jun 11, 2010 1:39 PM	nips
27	Jun 11, 2010 1:47 PM	holds for review 1 in the first phase and review 3 in the second phase. the other review in phase 1 was poor.
28	Jun 11, 2010 1:51 PM	ICML reviews are much better than SIGIR.
29	Jun 11, 2010 1:57 PM	NIPS, KDD ... machine learning, data mining
30	Jun 11, 2010 1:58 PM	NIPS
31	Jun 11, 2010 2:03 PM	AAAI, CogSci
32	Jun 11, 2010 2:31 PM	I guess we were simply unlucky this year.
33	Jun 11, 2010 2:51 PM	Nips
34	Jun 11, 2010 2:58 PM	nips
35	Jun 11, 2010 3:38 PM	My (rejected) paper was about manifold learning which fits ICML well. My home conference is CVPR. I think that papers that fit the core areas of CVPR usually receive more nuanced reviews.
36	Jun 11, 2010 4:38 PM	NIPS, MACHINE LEARNING
37	Jun 11, 2010 4:56 PM	I'm comparing against NIPS, ECML and AAAI in the area of reinforcement learning
38	Jun 11, 2010 6:15 PM	nips (about the same), ecml (better), iconip (better)
39	Jun 11, 2010 8:44 PM	Theory, STOC/SODA etc.
40	Jun 12, 2010 12:27 AM	AAAI, NIPS
41	Jun 12, 2010 3:57 AM	NIPS, CVPR, AISTATS
42	Jun 12, 2010 10:57 AM	NIPS: machine learning, CVPR/ICCV/ECCV: computer vision
43	Jun 12, 2010 11:14 AM	NIPS, AiSTAT
44	Jun 12, 2010 1:41 PM	SDM, KDD, ECML
45	Jun 12, 2010 2:39 PM	NIPS
46	Jun 12, 2010 11:27 PM	The second-phase reviews have been consistently poor, and this is frustrating because there is no associated response period. Otherwise "about the same".
47	Jun 13, 2010 1:15 AM	Can't say. This is my first Machine Learning submission.
48	Jun 13, 2010 5:27 PM	KDD, ICDM, classification, online learning
49	Jun 14, 2010 4:34 AM	AISTATS, NIPS; kernel methods sparse methods, supervised learning.
50	Jun 14, 2010 3:47 PM	i hoped the icml reviews were substantially better, unfortunately they were not
51	Jun 15, 2010 9:25 AM	NIPS
52	Jun 15, 2010 2:23 PM	In my personal experience, ICDM has typically offered more helpful feedback and more reasoned reviews than ICML.
53	Jun 19, 2010 5:18 AM	NIPS,IJCNN. The subject area is the neural networks.
54	Jun 25, 2010 6:44 AM	NIPS, KDD, SDM, ICDM
55	Jun 25, 2010 8:30 AM	Depends on the conference, in some of them they are about the same level, and in some others they are worse