Automation of distribution of students between graduate supervisors with application of two-sided matching model

Abstract

The article deals with an approach for modeling and software support of distribution of students between graduate supervisors at large graduate department. The approach is based on the stable matching problem and the Gale-Shapley deferred acceptance algorithm, and takes into account both students and supervisors’ preferences. The formalized description of distribution model is given, and the results of its practical verification are described. The advantages and disadvantages of the proposed approach are discussed, and the problem of preferences manipulation of graduate supervisors are examined. The architecture of the distribution support software system is presented, and some features of its implementation as a Web-service within the complex information system of the graduate department are described.

Keywords

• Graduate work
• two-sided matching
• stable matching
• deferred acceptance algorithm
• preferences manipulation