Abstract:
|
Peer review in conferences and grant panels is competitive, meaning that only a fixed number or fraction of papers or grants will be accepted. As a consequence, peer review suffers from dishonest behavior where reviewers manipulate the reviews they provide in order to improve the ranking of their own papers or proposals. In this talk, we will discuss methods to detect and mitigate such dishonest behavior. The methods are supplemented with strong theoretical guarantees and empirical evaluations.
|