nteractive Decision Support System for Planned Student Scheduling and Incremental Changes in Large Scale Timetabling

Abstract

A student joins an institution to earn some qualification. For this purpose, he/she has to successfully complete a set of courses in a span of few terms. Some of the students fail in some of the courses studied in a term. Such courses are called backlog courses. Backlog courses have to be taken up in any of the successive terms possible only if these are offered to junior class students in one or more sections. The planning needed to help students in deciding which courses can be taken with which of the section of junior class students, forms the core of present thesis. Besides this, the thesis also covers incremental changes to be made in a working time table once the whole process has stabilized. This thesis consists of six chapters. Chapter One is introductory in nature. In this chapter, we briefly discuss the issues of backlog courses while preparing timetable. Two approaches generally used to handle this problem and some important terminologies are also discussed in this chapter. A brief survey of the literature available on the subject, and summary of the work presented in the succeeding chapters of the thesis, also appears in this chapter. The second chapter describes the model for planning of student scheduling. Planning of student scheduling involves the linking of courses that the students will be attending with students of junior classes while they are free. This scenario is solved by a method developed to find out backlog courses which can be pooled together in same time slot along with keeping free the sections of students where regular courses are held. This process is termed as planning for student scheduling whose output can be considered as a set of constraints to be used in building the time table. The Hungarian method of assignment model has been repeatedly used to form links further required as an input for scheduling of the timetable. A core model has been designed using Dynamic Programming approach spread across five stages where each stage is referring to one year of study.