[Corpora-List] Against the reviewer mediation stage

[Miles Osborne]

As someone who recently been both a reviewer AND an area chair (and
also often reviews gigantic numbers of papers), one of the good things
about the conference reviewing procedure is the reconcilliation (and
truth?) stage.  As a reviewer, I'm usually in agreement with my
co-reviewers *but not always*.  And as a chair, this helps me manage
the large number of reviews.

In my experience, there is generally little reconcilliation that
requires enormous amounts of effort.  Usually it can be done by reply;
 and if not, a re-read of the paper does not take time at all.

Note:  the key to reviewing not being a crushing duty is successful
bidding.  This makes all the difference.  I've stopped reviewing for
ECML as they recently moved over to non-bidding.

Now, one import from the Machine Learning community which is of
questionable utility is Author Feedback:  it is not clear to me what
the role of this is, nor how useful it is.  This can draw-out the
reviewing process even longer as there is yet another stage to deal
with.

(On a nice note, a paper we wrote  last year at a ML venue actually
had an increase in a score after we as authors responded, so I guess
it  is not all bad).

Miles

-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora