An adaptive strategy for peer review

Peer review has been widely used in multiple contexts ranging from childhood education to academic research. Numerous documents describe different aspects of the reviewing process. Interestingly, barely any discussion is found about how to better match peer assessors and assesses. In this paper an application of peer review in an educational environment is presented. In such context, it is considered as a pedagogical tool to improve the overall learning experience. It is shown that by carefully assigning reviewers to authors, peer review can be oriented to maximize the benefits of the process. Normally, the mapping between authors and reviewers is simplified to a blinded, random association. The proposed adaptive strategy suggests an effective method for assigning reviewers and authors depending on user profiles and based on fuzzy classification techniques. The strategy requires a set of prototypes (typical author-reviewer pairs) and a set of corresponding weights specifying the interest of each prototype. Genetic algorithms are used to obtain a solution to optimize the assignments of reviewers to authors.