Models, solution methods and threshold behaviour for the teaching space allocation problem

Beyrouthy, Camille (2008) Models, solution methods and threshold behaviour for the teaching space allocation problem. PhD thesis, University of Nottingham.

[thumbnail of thesis.pdf]
Preview
PDF - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (4MB) | Preview

Abstract

Universities have to manage their teaching space, and plan future needs. Their efforts are frequently hampered by, capital and maintenance costs, on one hand, pedagogical and teaching services on the other. The efficiency of space usage, can be measured by the utilisation: the percentage of available seat-hours actually used. The observed utilisation, in many institutions, is unacceptably low, and this provides our main underlying motivation: To address and assess some of the major factors that affect teaching space usage in the hope of improving it in practise. Also, when performing space management, managers operate within a limited number and capacity of lecture theatres, tutorial rooms, etc. Hence, some teaching activities require splitting into different groups. For example, lectures being too large to fit in any one room and seminars/tutorials being taught in small groups for good teaching practise. This thesis forms the cornerstone of ongoing research to illuminate issues stemming from poorly utilised space and studies the nature of constraints that underlies

those low levels of utilisation. We give quantitative evidence that constraints related to timetabling are major players in pushing down utilisation levels and also, devise "Dynamic Splitting" algorithms to illustrate the effects of splitting on utilisation levels. We showed the existence of threshold between phases where splitting and allocation is "always possible" to ones where "it's never possible", hence, introducing a practical application of Phase Transition to space planning and management. We have also worked on the long-term planning aspect of teaching space and proposed methods to improve the future expected utilisation.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Burke, Edmund
Keywords: AI Models, threshold behavior, Integer programing, Space Allocation, timetabling.
Subjects: Q Science > QA Mathematics
Faculties/Schools: UK Campuses > Faculty of Science > School of Computer Science
Item ID: 10576
Depositing User: EP, Services
Date Deposited: 24 Oct 2008
Last Modified: 14 Oct 2017 20:43
URI: https://eprints.nottingham.ac.uk/id/eprint/10576

Actions (Archive Staff Only)

Edit View Edit View