Per i Big Data nella ricerca valutativa: una proposta operativa per la gestione del matching prodotto-revisore nella VQR.

Moving from some critical issues that characterized the management of the product-reviewer matching in the VQR 2011-2014 this article proposes an operational solution that allows improving the consistency of the attribution process, by bringing the gap between the reviewer's skills and the evaluand. This is achieved through the use of text analysis techniques applied to big corpora, made by the publications/abstracts of the actors involved. The paper is structured as follows: an introduction about the use of Big Data in the context of the Italian university system evaluation; then, the main problems faced in the last VQR exercise will be presented; finally, a textual analysis procedure, designed to provide univocal labels, both for the products under review and for those belonging to referees, will be presented in the last paragraph. [ABSTRACT FROM AUTHOR]