Paper Rating vs. Paper Ranking

John R. Douceur

Abstract

Within the computer-science community, submitted conference papers are typically evaluated by means of rating, in two respects: First, individual reviewers are asked to provide their evaluations of papers by assigning a rating to each paper’s overall quality. Second, program committees collectively rate each paper as being either worthy or unworthy of acceptance, according to the aggregate judgment of the committee members. This paper proposes an alternative approach to these two processes, based on rankings rather than ratings. It also presents experiences from employing rankings in PC discussions of a major CS conference.

Details

Publication typeArticle
Published inOperating Systems Review 43 (2)
PublisherAssociation for Computing Machinery, Inc.

Previous versions

John R. Douceur. Paper Rating vs. Paper Ranking, USENIX, 2008.

> Publications > Paper Rating vs. Paper Ranking