Project allocation using optimisation techniques : a method to allocate FYP projects for Nanyang Technological University

Abstract

In this report we present a method to allocate FYP projects for Nanyang Technological University. We model the FYP allocation problem as a propositional satisfiability problem. Once the FYP allocation problem is transformed, we use programs, so called SAT solvers, to solve the propositional satisfiability problems by finding a satisfying assignment. We first show how several requirements of allocating FYP projects can be stated as constraints. Next, these constraints will be written into a file using a Java program. Lastly, we use the SAT solvers to find a satisfying assignment and thereby allocations to all students. In this project, Conjunctive Normal Form Format is first used to formulate the problem. However, it is later found out that there are constraints that cannot be formulated using this format especially given the constraint that each lecturer shouldn’t be allocated students of CGPA more than a certain average. Even though this is an extra feature given by his professor, and that the deadline of the project is near, the student decided to give the Pseudo Boolean format a try and the results were shown to be optimum. Lastly, we will see how an actual NTU allocation data with 247 students can be solved with the CNF solver and the weighted MAXSAT solver.