Review Coherence
Coherence between
numerical scores and review content
INFOCOM 2014 will adopt a numerical score range of 1 to 5 for reviews. The TPC lead assigned to a given paper will have to check if the numerical scores are well justified by the written reviews (and also catch short/low quality reviews), and follow up with reviewers using EDAS. Since every TPC member is a lead for some papers, kindly use this opportunity to check the coherence of the scores/reviews you give to papers.
For example, if there is a score of 5, there has to be clear
enumeration of reasons why this paper deserves that rating.
(Remember that the scores are logarithmic!) After you have reviewed
all papers, and see the distribution of scores you have given,
please see if there is any bias in your scoring and if it is indeed
justified by the quality of the papers in your set.
Similarly, a rating of 2 or 1 may lead to a paper not making it to
the TPC meeting: so it is the reviewer's responsibility to give more
feedback for authors. Please note that irrespective of your scores,
if a paper has reviews that are not strong, the scores may not
matter.
