p**n
发帖数: 261 | 1 Our improvement over previous methods is from several times to 100 times faster.
Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
Yet none of them ever changd any word in their reviews, based on our feedback.
Sigmod reviewers are really pathetic! |
l*********r
发帖数: 674 | 2 岂止sigmod，现在很多会都是这样，reviewer就是为了据而据。现在要中一篇paper更
多是看你的运气，能不能同时碰到3个认真的reviewer。而且我感觉这种趋势这两年还
越来越明显，最近两年我的好几篇paper都是被无厘头地据了，前几年都没有碰到过这
么多恶心的事情。以前还想毕业后进acdemic，现在义无反顾地投向工业界。
faster.
too good to be true".
insignificant, considering the simplicity of the previous method".
feedback.
【在 p**n 的大作中提到】
: Our improvement over previous methods is from several times to 100 times faster.
: Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
: Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
: Yet none of them ever changd any word in their reviews, based on our feedback.
: Sigmod reviewers are really pathetic!
|
p**n
发帖数: 261 | 3 I agree. I have seen quite a few bad reviewers of top conferences. Some of
them do not even understand what the paper talks about but they do not
hesitate to give terribly wrong comments.
I suspect some professors just ask their students to do reviews.
【在 l*********r 的大作中提到】
: 岂止sigmod，现在很多会都是这样，reviewer就是为了据而据。现在要中一篇paper更
: 多是看你的运气，能不能同时碰到3个认真的reviewer。而且我感觉这种趋势这两年还
: 越来越明显，最近两年我的好几篇paper都是被无厘头地据了，前几年都没有碰到过这
: 么多恶心的事情。以前还想毕业后进acdemic，现在义无反顾地投向工业界。
:
: faster.
: too good to be true".
: insignificant, considering the simplicity of the previous method".
: feedback.
|
a********y
发帖数: 386 | 4 How did you mention you 100 time improvement...?
faster.
too good to be true".
insignificant, considering the simplicity of the previous method".
feedback.
【在 p**n 的大作中提到】
: Our improvement over previous methods is from several times to 100 times faster.
: Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
: Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
: Yet none of them ever changd any word in their reviews, based on our feedback.
: Sigmod reviewers are really pathetic!
|
y***u
发帖数: 101 | 5 Yes we all have had reviewers like this. And indeed, many reviewers just
want
to kill your paper. So from my experience to get a paper into sigmod you
need:
1. Luck
2. Prove the superiority of your new method, instead of just showing the
experimental results. Experimental results can be easily attacked: the
data is not right, you didn't implement the old method right, you didn't
implement yours right, bla bla bla
3. If you can't do 2, give very good intuitive reasons why your method is
【在 p**n 的大作中提到】
: Our improvement over previous methods is from several times to 100 times faster.
: Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
: Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
: Yet none of them ever changd any word in their reviews, based on our feedback.
: Sigmod reviewers are really pathetic!
|
t*s
发帖数: 1504 | 6 is sigmod double blind
faster.
too good to be true".
insignificant, considering the simplicity of the previous method".
feedback.
【在 p**n 的大作中提到】
: Our improvement over previous methods is from several times to 100 times faster.
: Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
: Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
: Yet none of them ever changd any word in their reviews, based on our feedback.
: Sigmod reviewers are really pathetic!
|
t*s
发帖数: 1504 | 7 学生做review还认真些
大佬自己review有的只看名字和abstract
【在 p**n 的大作中提到】
: I agree. I have seen quite a few bad reviewers of top conferences. Some of
: them do not even understand what the paper talks about but they do not
: hesitate to give terribly wrong comments.
: I suspect some professors just ask their students to do reviews.
|
l******e
发帖数: 470 | 8 不认真的学生多了去了
【在 t*s 的大作中提到】
: 学生做review还认真些
: 大佬自己review有的只看名字和abstract
|
l******e
发帖数: 470 | 9 现在的趋势是paper审的越来越不认真。对写作的要求越来越高，而technical
comtribution和conceptual contribution却不那么重要。reviewer 不是努力找优点来
accept，而是努力找缺点来reject。一篇有solid contribution的paper可能因为几个
typo被拒，而一篇很一般的paper却因为写的不错而收了。和theory community正好相
反。
岂止sigmod，现在很多会都是这样，reviewer就是为了据而据。现在要中一篇paper更
多是看你的运气，能不能同时碰到3个认真的reviewer。而且我感觉这种趋势这两年还
越来越明显，最近两年我的好几篇paper都是被无厘头地据了，前几年都没有碰到过这
么多恶心的事情。以前还想毕业后进acdemic，现在义无反顾地投向工业界。
faster.
too good to be true".
insignificant, considering the simplicity of the previous method".
feedback.
【在 l*********r 的大作中提到】
: 岂止sigmod，现在很多会都是这样，reviewer就是为了据而据。现在要中一篇paper更
: 多是看你的运气，能不能同时碰到3个认真的reviewer。而且我感觉这种趋势这两年还
: 越来越明显，最近两年我的好几篇paper都是被无厘头地据了，前几年都没有碰到过这
: 么多恶心的事情。以前还想毕业后进acdemic，现在义无反顾地投向工业界。
:
: faster.
: too good to be true".
: insignificant, considering the simplicity of the previous method".
: feedback.
|
r**m
发帖数: 163 | 10 是double blind,真希望所有都double blind，可惜KDD就不是
【在 t*s 的大作中提到】
: is sigmod double blind
:
: faster.
: too good to be true".
: insignificant, considering the simplicity of the previous method".
: feedback.
|
|
|
l****a
发帖数: 192 | 11 Similar situation happened here.
From reviewer 1, one of the strong points is "the technique generalizes well
to other DB systems."
From reviewer 2, one of the weak points is "the technique does not
generalize; it is only evaluated on one DB system." Actually, we did
evaluation on both MySQL and Postgres. How poor the review is! "why did not
you do evaluation on commercial DBs?" Who paid us for buying commercial DBs
? More questions might come: why DB2? why not oracle?
I agree some reviewers ju |
x*j
发帖数: 271 | 12 Same here. Too hard to get a good
reviewer. It seemed to me that in general VLDB's review is better than
SIGMOD's. |
h****o
发帖数: 2455 | 13 reviewer也需要被理解。现在的趋势是太多为了发paper而发paper的了，reviewer也不
可能真花时间去验证你的结果，所以有时候看着不顺眼就找个理由给拒掉。 |
v****s
发帖数: 1112 | 14 are you a good reviewer if u r reviewing a paper? |
W******g
发帖数: 887 | |
z*y
发帖数: 1311 | 16 Just ask, is the old idea presented in top conference or not?
faster.
too good to be true".
insignificant, considering the simplicity of the previous method".
feedback.
【在 p**n 的大作中提到】
: Our improvement over previous methods is from several times to 100 times faster.
: Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
: Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
: Yet none of them ever changd any word in their reviews, based on our feedback.
: Sigmod reviewers are really pathetic!
|
n***l
发帖数: 1468 | |
v******d
发帖数: 1322 | 18 Oracle is free for non commercial use. Many universities also have access to
most commercial dbs.
well
not
DBs
【在 l****a 的大作中提到】
: Similar situation happened here.
: From reviewer 1, one of the strong points is "the technique generalizes well
: to other DB systems."
: From reviewer 2, one of the weak points is "the technique does not
: generalize; it is only evaluated on one DB system." Actually, we did
: evaluation on both MySQL and Postgres. How poor the review is! "why did not
: you do evaluation on commercial DBs?" Who paid us for buying commercial DBs
: ? More questions might come: why DB2? why not oracle?
: I agree some reviewers ju
|
R********n
发帖数: 519 | 19 说得很好，现在大都偏应用或者很应用的领域都有这个问题，大多数情况下门槛不会太
高，
做得人增加，paper投稿量增加，于是同一个大的idea，很多类似文章出现。能中的就
是写
得好的，还往往不是真正solid的paper。
经常能感觉顶级conf上一些的paper，从头到尾粗看无懈可击，introduction,model,
experiments,figure,equations，都很漂亮。真正看懂了之后，确发现没什么创新，只是
A + B 或者 A->A'。
我觉得这就是形式主义的问题。根源就是做得人太多了和体制问题，导致现实和理想的
research相距太远
【在 l******e 的大作中提到】
: 现在的趋势是paper审的越来越不认真。对写作的要求越来越高，而technical
: comtribution和conceptual contribution却不那么重要。reviewer 不是努力找优点来
: accept，而是努力找缺点来reject。一篇有solid contribution的paper可能因为几个
: typo被拒，而一篇很一般的paper却因为写的不错而收了。和theory community正好相
: 反。
:
: 岂止sigmod，现在很多会都是这样，reviewer就是为了据而据。现在要中一篇paper更
: 多是看你的运气，能不能同时碰到3个认真的reviewer。而且我感觉这种趋势这两年还
: 越来越明显，最近两年我的好几篇paper都是被无厘头地据了，前几年都没有碰到过这
: 么多恶心的事情。以前还想毕业后进acdemic，现在义无反顾地投向工业界。
|
w****8
发帖数: 149 | 20 A+B越来越普遍了。。。
如何定义solid呢？
只是
【在 R********n 的大作中提到】
: 说得很好，现在大都偏应用或者很应用的领域都有这个问题，大多数情况下门槛不会太
: 高，
: 做得人增加，paper投稿量增加，于是同一个大的idea，很多类似文章出现。能中的就
: 是写
: 得好的，还往往不是真正solid的paper。
: 经常能感觉顶级conf上一些的paper，从头到尾粗看无懈可击，introduction,model,
: experiments,figure,equations，都很漂亮。真正看懂了之后，确发现没什么创新，只是
: A + B 或者 A->A'。
: 我觉得这就是形式主义的问题。根源就是做得人太多了和体制问题，导致现实和理想的
: research相距太远
|
|
|
s****d
发帖数: 56 | 21 In sigmod, usually there are few non-rejectable paper(stronger than 3
accepts), but a lot of boundary paper(not all reviewers vote an accept).
Whether a boundary paper can be accepted or not depends on which group/
professor/co-authors the paper comes from. The clique of sigmod is
relatively tight. But VLDB/ICDE have a broader community.
Normally reviewers do not change their comments after reading author
feedback. If you have a feedback, it often means reviewers have conflict
comments and your
【在 p**n 的大作中提到】
: Our improvement over previous methods is from several times to 100 times faster.
: Reviewer A's main reason for rejection is like "The result is spurious, too good to be true".
: Reviewer B's main reason for rejection is like "The improvement is insignificant, considering the simplicity of the previous method".
: Yet none of them ever changd any word in their reviews, based on our feedback.
: Sigmod reviewers are really pathetic!
|
y***u
发帖数: 101 | 22
This is interesting. Do you have any statistics to back up this claim?
【在 s****d 的大作中提到】
: In sigmod, usually there are few non-rejectable paper(stronger than 3
: accepts), but a lot of boundary paper(not all reviewers vote an accept).
: Whether a boundary paper can be accepted or not depends on which group/
: professor/co-authors the paper comes from. The clique of sigmod is
: relatively tight. But VLDB/ICDE have a broader community.
: Normally reviewers do not change their comments after reading author
: feedback. If you have a feedback, it often means reviewers have conflict
: comments and your
|
p*******n
发帖数: 4824 | |
