Student project allocation using integer programming

Abstract

Abstract-The allocation of projects to students is a generic problem in many universities within the U.K. and elsewhere, not only in engineering but also in various other disciplines. This paper defines the student project allocation problem explicitly by an objective function and a number of constraints. Two integer program models are presented, the first of which is a dynamic program. A general purpose solver is used to solve the models, and the input files are included in the Appendix. The models are computationally efficient and easily solved on a PC. Important issues in interpreting the model outputs are highlighted. As with any optimization problem it is possible for constraints to be too tight to permit any feasible solution. Application of the models is demonstrated by using data from the Department of Civil and Environmental Engineering, University of Southampton, for the academic year 2001-2002. The model has been used successfully to allocate Individual Project and Group Project to students and is likely to become the defacto method of allocation of projects in the future. This paper demonstrates how operations research techniques used widely in optimizing use of resources can be applied in education.Index Terms-Allocation model, group project, individual project, integer programing, University of Southampton. NOMENCLATUREBinary integer that assumes a value of 1 if a student is of a discipline . Binary integer that assumes a value of 1 if a discipline is suited to project and 0 otherwise. Binary integer that assumes a value of 1 if a student is of a discipline that is a subset of the disciplines required for project and 0 otherwise. Number of students allocated any project . Index representing a student and can assume a value . Index representing a suggested project and can assume a value . Integer representing a discipline and can assume a value . Late penalty factor. Maximum number of students permitted on any project . Total number of disciplines. Minimum number of students permitted on any project . Total number of suggested projects. Total number of students. Binary variable which assumes a value of 1 if a project has been allocated and 0 otherwise. Binary integer that assumes the value of 1 if the project is selected by the student , and 0 otherwise. Binary decision variable for the th student and th project.Objective function variable for model 1a. Objective function for model 1b.Large positive number. Total number of projects tutored by a member of staff. I. BACKGROUNDA LLOCATION of projects to students as part of a degree course is common to most, if not all, universities. However, the exact requirements and procedures differ widely. Any project allocation model/algorithm must take into account the specific conditions under which it is being applied. For the purpose of this paper, the Department of Civil and Environmental Engineering, University of Southampton, is taken as a case study, and its particular conditions for project allocation are explained at some length. However, the models pres...

A Multi-Objective Approach for the Project Allocation Problem

“…By doing so, it is expected that the whole system will now be fairer to everyone. [3] presents two solutions for the project allocation problem. In the first model, every student is allocated one project at random so that each supervisor has at least one project to supervise.…”

Student-Project Allocation with preferences over Projects

“…As far as we are aware, spa-p is the first matching problem of this type to be considered in the literature. The previous formulations of spa to have been considered either do not permit lecturer preferences [3,8,10,12] (so stability is not relevant in these contexts) or involve lecturer preferences over students [1,2,4,9,13]. …”

Course Opening, Assignment and Timetabling with Student Preferences

“…Methods for solving the SPA problem are, among others, building an integer program and solving it with a solver Anwar and Bahaj (2003), defining a linear program for some particular cases Saber and Ghosh (2001), or developing specific heuristics like a genetic algorithm Harper et al (2005). Manlove and O'Malley Manlove and O'Malley (2008) show the complexity and give an approximation algorithm for a generalized SPA problem in which both the students and the lecturers have preferences over the projects.…”