Professional Documents
Culture Documents
Examination Scheduling System Based On Quadratic Assignment: October 2014
Examination Scheduling System Based On Quadratic Assignment: October 2014
net/publication/266477253
CITATIONS READS
2 609
6 authors, including:
Some of the authors of this publication are also working on these related projects:
Recent advances in energy efficient-QoS aware MAC protocols for wireless sensor network View project
All content following this page was uploaded by Muktar Danlami on 14 June 2019.
Danlami Muktar, Fadhilah Ahmad, Zarina Mohammad, Hasni Hassan, Aznida Hayati Zakaria
{si0497@putra, fad@, zarina@, hasni@, aznida@}.unisza.edu.my
Faculty of Informatics and Computing (FIC),
Sultan Zainal Abidin University (UniSZA),
Tembila Campus, 22200 Besut, Terengganu, Malaysia
a Decision Support System ensuring the Timetabling Problem (ETP), a large number of
following constraints are satisfied: researchers are interested in the applications of
No student sits for more than one heuristics to solve the problem.
examination at the same time. The table below in Table 2.1 provides related
No student should take more than one work done by different authors using Algorithm
course in within a day.
Achieve full capacity utilization of Table 2.1 Related works using Optimization
building facilities and ensure capacity of Algorithms
individual rooms is not exceeded at any No Author/Date Algorithm
time throughout the examination 1 Desroches, Larporte, Travelling
session. and Rousseau (1978) Salesman
More than one courses can be and White and Chan problem (TSP)
conducted in the same room (1979)[4][9]
Duration of examination period are not 2 Aldy Gunawan Hybrid
exceeded. 2008[5]
Lecturers are considered when 3 Mohd Zaki Bin Mohd Genetic
assigning invigilating sessions to them. Salikon 2005[6]
One of the invigilators is the lecturer 4 Nelishia Pillay Hyper-Heuristic
who teaches the subject 2010[7]
5 Nasser R. Sabar, Graph Coloring
2. PROBLEM STATEMENT Masri Ayob Rong Qu, Heuristic
Graham Kendall
An improper examination timetable may result 2011[8]
in students' poor performance as it may not let
them enough preparation between two The earlier studies by [4][9]adopt the
sequential exams. In addition, there are many Travelling Salesman problem (TSP) approach
exams to be scheduled, large number of with the use of list generation to eliminate the
students who have taken different courses, the conflicts of when one or more students are
limited number of rooms or exam halls are assigned to sit for two or more papers in the
available, and some constraints such as no same session. These methods do not consider
conflict in a single student's exams make it very more complex conflicts.
difficult to schedule experimentally. The A related work by [8] investigated a
scheduling problem created by these set of new graph coloring constructive hyper-heuristic
circumstances clearly poses an interesting for solving examination timetabling problems
intellectual challenge. This study explores this using hierarchical hybridizations of four low
challenge and proposes solution to this level graph coloring heuristics. The solution
problem. lists all examinations by the level of difficulty
of assignment and ensures the most difficult
2.1 Related Theories exam to be scheduled is scheduled first. This
method was considered simple and yet efficient
Examination timetabling is a well-researched approach but suffers setbacks in handling
and important educational timetabling problem higher level conflicts in examination
[3]. It has been studied many times in other scheduling. Graph coloring has the limitation
institutions since the mid-1960s. Although which neglects hard constraints such as
there are some novel research that focused on students cannot take two examinations
the modeling and formulation of Examination consecutively by formulating the basic
The model assigned the invigilators using Due to wide assortment of applications, The
arbitrarily assigned ratio scale numbers for the QAP has been extensively used to formulate
exam weights. other practical problems such as economic
Recently, [2], developed a new binary model problems, Scheduling problems, Timetabling
for examination time tabling problem, the problems, Facility layout problems,
model was presented in two directional novelty.
The first one is that a course can be offered
more than once in a semester. If a course is 3.2 Mathematical Formulation
requested by a few students, then it is enough to
be offered once. If the number of students Taking E exams need to be scheduled into a
requesting a course is more than the maximum limited number of T time slots and limited
number of students who are allowed to attend a number of R room and capacity, taking into
single class, then the course is multi-offered. account all examination must be scheduled
The second novelty is that sharing a room for within the time frame and the rooms capacities
two simultaneous exams is allowed. Also, the and the restrictions that students cannot take
model considers some hard and soft constraints, more than one examination at the same time
and the objective function is set in such a way among other constraints.
that soft constraints are satisfied as much as
possible N= {1, 2, 3 … n}
n T : E R , the set of all permutations.
3. PROPOSED FRAMEWORK tij T , is the time slot for an exam to be
Timetabling problems have attracted the assigned to a room
n n
continuous interest of researchers mainly
because they provide the opportunity of testing
min ij tij
i 1 j 1
combinatorial solution methods in formulations
that represent difficult practical problems i = 1, 2, 3 … n
(Dimopolou, Miliotis, 2000). With anticipated j = 1, 2, 3 ... (n-1)
further increase in enrolment and number of
courses, greater flexibility in choice of courses, Is represent the assignment of exams to
and the need to keep the overall length of the specific time slot and different room, so as to
final examination period short, a more prevent conflict. The conflicts degree of each
systematic analytical procedure is necessary. examination is minimized when exams i are
assigned to room j at a specific time and is
3.1 Quadratic Assignment Problem denoted by ij tij where
Exam E1,4 and 9 scheduled to take place in scheduled for T3 and Exam E5, 6 and 10 for the
day1 session1 which is time slot1(T1), Exam second day last session which is T4
E3 and 7 scheduled at T2, Exam E2 and 8
4. CONCLUSION AND
RECOMMENDATION
Proposed Algorithm
In this study, we highlighted some weaknesses
Here, we propose an algorithm based on and strength of the related studies on
heuristic. This solution identifies possible examination scheduling. The proposed
constraints and groups them as hard and soft examination timetabling problem was
constraints. The algorithm ensure the solution formulated based a Quadratic Assignment
is improved if need be until an optimal solution Problem (QAP). Since the number of
is achieved and all constraints are satisfied. examinations can be greater than the number of
The algorithm process is described below and time periods and in a quest to minimize all
also presented in the diagram in Fig 3.4: possible conflicts, the objective function is set
in such a way that, hard constraint satisfaction
1. Identify all possible constraints and is maximized. The future work of this research
group them in the order of hard to soft is to implement an examination scheduling
constraints system which will allow simulation and
2. Select the hardest constraint and apply evaluation of various examination data and
solution provide results that will help examination
3. Check if constraint is satisfied timetable planner achieve quality and valuable
4. If constraints are satisfied, then repeat scheduling process.
process for all other constraint
otherwise improve solution.
5. If solution is final solution, then apply 5. Acknowledgements
to constraint until all constraints are
satisfied. The authors would like to extend appreciation
of the support given by University Sultan
Zainal Abidin (UNISZA) and Kano State
Government. The authors are grateful to Mr
Aiyerin Taiwo, Mr. Saleem and Malam Aminu
for their contribution and support to the work.
6. REFERENCES
[7] Nelishia Pillay, “An Empirical Study into the [14] Andrew L, Ang J. C, Ho W. K, Oon W C. “A
Structure of Heuristic Combinations in an Evolutionary Campus-wide University Examination Timetabling
Algorithm HyperHeuristic for the Examination Application”. American Association for Artificial
Timetabling Problem”, SAICSIT ’10, October 11–13, Intelligence. pp 2-3 May, 2000.
2010, Bela Bela, South Africa.
[15] Tomas M. “Real-life Examination Timetabling.
[8] Nasser R.Sabar·Masri Ayob·RongQu· Graham Space Management and Academic Scheduling” - Purdue
Kendall “A graph coloring constructive hyper-heuristic University pp 1-2. June, 2013
for examination timetabling problems” Springer Science
+ Business Media, LLC 2011 [16] Thin-Yin L, Wee-Yong Y. “A Hierarchical
Decision Support System for University Examination
Scheduling” pp 4-10. SEPT, 1990
[9] White G M and Chan P-W (1979), "Towards the
Construction of Optimal Examination Schedules," [17] Barry M, Paul M. “A New Model for
INFOR 17, 219-229 Automated Examination Timetabling”. pp 3-5 April,
2008.
[10] Dimopoulou, M., & Miliotis, P. (2001).
Implementation of a university course and examination
timetabling system. European Journal of Operational
Research,130, 202–213