CS SR
14 - 15 March 2009
0 8’ 0 9
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
EXAMINATION INVIGILATION SCHEDULING SYSTEM IN OPTIMISING LECTURERS’ PREFERENCE 1
2
3
Associate Professor Ong Mooi Lian , Liew Lee Hung and Jacqueline Sim 1
Faculty of Civil Engineering, Universiti Teknologi MARA, Kota Samarahan, SARAWAK Faculty of Information Technology & Quantitative Sciences, Universiti Teknologi MARA, Kota Samarahan, SARAWAK 3 Academy of Language Studies, Universiti Teknologi MARA, Kota Samarahan, SARAWAK
[email protected] [email protected] [email protected]
2
ABSTRACT The examination invigilation scheduling system is purposed to facilitate the preparation of examination invigilation schedules so as to reduce manual involvement and the amount of time taken to prepare examination invigilation schedules. This system is made possible with the creation of a database that collects information pertaining to the examination invigilation schedules such as lecturers’ preferences of invigilation dates and time and their constraints. Besides, the system also allows lecturers to provide feedback and any other relevant information to the invigilation scheduling committee. On top of that, being computer-aided and easily accessible by lecturers online, the lecturers can utilise the system at their own convenience. When the invigilation schedule is ready, the lecturers can obtain a summary of their individual invigilation duty, and view the schedule as well as all the relevant invigilation rules and regulations within and outside the office through the Internet. This system benefits especially the Academic Administration Office of institutions of higher learning because it enables them to produce comprehensive examination invigilation schedule. This automated way of capturing, optimising and disseminating invigilation information paves the way for the development of a systematic approach in preparing examination invigilation schedules. With that, suitable problem-solving support systems which will aid the process of computerising examination invigilation scheduling system can be identified. This examination invigilation scheduling system which has been found to be successful in UiTM Sarawak since 1995 has been going through continuous improvement.
Keywords: invigilation, scheduling, database, online, preferences
1. INTRODUCTION
Final examination invigilation of students at Universiti Teknologi MARA Negeri Sarawak Samarahan Campus (UiTM Sarawak) involves all the lecturing staff who follows a set of invigilation schedule that is prepared manually by fellow lecturers in the university. Preparing the examination invigilation schedule has always been in itself a challenging task as the invigilation schedule committee has to take into consideration numerous factors and lecturers’ constraints such as getting the invigilation schedules ready within a limited time frame, ensuring the invigilation duties assigned to the lecturers do not disrupt their marking and that the lecturers do not invigilate their own subjects. Prior to the present invigilation scheduling committee taking over the responsibility of preparing the invigilation schedules, invigilation duties were assigned randomly and, hence, there were a lot of mutual swapping amongst the lecturers resulting in confusion, misunderstandings and complaints on uneven duty distribution. Thus, when the present invigilation scheduling committee took over the task of invigilation Paper number: [1897541]
CS SR
0 8’ 0 9
14 - 15 March 2009
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
scheduling in 2004, they saw the need for a systematic and yet innovative approach to produce invigilation invigilation schedules that could minimise errors and simultaneously allowed lecturers to request for their preferred invigilation dates and time. The system worked well but eventually with the increasing amount of data, provided by the growing population of lecturers, on the time and dates of their examination papers coupled with their preferred invigilation dates, the committee which had been manually transferring data into the computer realised an urgent need for a computer-aided system that could minimise manual involvement and processing time in data handling. Therefore, this system which is the examination invigilation scheduling system (ISS) aims to reduce manual involvement and time taken by the scheduling committee in developing a systematic approach in the preparation of invigilation schedules that can capture the lecturers’ preference online. With this in mind, the committee creates a centralised database for collecting information, and identifies suitable software systems which facilitates the data processing process and supports the ISS on-line. Thus, the examination invigilation schedules can be optimised based on the preferences of lecturers and the constraints faced. The development of this system is significant as it is capable of minimising the involvement and time spent by the committee on the preparation of invigilation schedules. Besides, it ensures consistency, reliability and continuity in the invigilation schedules produced in future semesters. The management of UiTM Sarawak through the Final Examination Unit in the Academic Administration Office will then be able to have full control over the final examination. The lecturers’ preferences will be optimised with the constraints controlled, and the number of alterations made at the Examination Operation Room to the final list of invigilators will be greatly reduced 2. LITERATURE REVIEW
Twenty-nine per cent of the universities surveyed by Burke et al. (1996) agree that the task of scheduling invigilators is a major problem. However, he adds that some universities have examination schedules that are so well-planned that they can be reused with “only a small alteration” (Burke et al., 1996: 7). Burke stresses that the quality of any schedule lies in its “usability and requirements of those who are subject to it” (Burke et al., 1996: 9). This type of system has to be flexible enough to successfully create schedules that not only allow room for any minor changes to be made to the master schedule but have also considered all the various constraints. According to Hishammuddin et al. (2004), there are various methodologies on examination scheduling which include constraint programming, graph colouring and metaheuristic approaches such as genetic algorithms, tabu search, simulated annealing, the great deluge algorithm and hybridised hybridised methods. Since the introduction introduction of the fuzzy methodologies methodologies into the scene of scheduling, it has experienced much success especially in maintenance scheduling (Dahal, Aldridge and McDonald, 1999), manpower allocation, aircrew roistering (Teodorovic et al.,1998), and driver scheduling (Li and Kwan, 2003). Hishammuddin and his team of researchers further investigate a fuzzy expert system that is used simultaneously with the multiple multiple ordering criteria in trying to give a better representation of the challenges encountered in examination scheduling. Their study led them to discover that the “exploration of different methodologies to achieve multi-criteria orderings may be beneficial” (Hishammuddin et al., 2004: 13). Burke et al. (2000) reiterates that various types of constraints which cannot be made possible using a single objective function can be handled by the multi-criteria approach. The system in this project utilises the multi-criteria modeling in its ISS. Burke et al.’s (1996) findings further reveal that automated timetabling packages are capable of reducing greatly the time spent on preparing examination schedules. This is supported by Junginger (1986) who discovers that using automated systems to prepare school timetables can save as much as 75 per cent of the time taken to plan the timetable using other non-automated systems.
Paper number: [1897541]
CS SR
14 - 15 March 2009
0 8’ 0 9
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
Thus, the invigilation scheduling committee of UiTM Sarawak will benefit greatly with a computerised ISS. The committee has since adopted an approach that can produce schedules that optimises both the lecturers’ preferences and the usage of manpower through the use of computer-aided applications such as MS Excel spreadsheets, Dream Weaver and MS Access. This approach, tested and proven over the past semesters, has produced schedules which are fairer in terms of the distribution of lecturers’ invigilation duties. Besides, the committee’s manual involvement in scheduling has been drastically reduced with the introduction of these computer-aided applications in the scheduling process. The committee has also initiated a platform of communication using computer-aided on-line system to allow lecturers to: i. ii. iii. iv.
view the examination time table; choose their preferred invigilation time slots; specify the examination date and time of their own subjects; and view their individual invigilation schedule and final examination/invigilation examination/invigilation timetable.
3. METHODOLOGY
ISS comprises the development of an online system for data collection and dissemination of assigned invigilation duties, and a systematic approach in assigning duties. There are three important modules to the online ISS, and each of them is used in different stages of the scheduling process. Figure 1 shows the data flow diagram of the ISS.
Figure 1: A data flow diagram (level 0) of the ISS online system Paper number: [1897541]
CS SR
0 8’ 0 9
14 - 15 March 2009
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
The three online ISS modules are: 3.1 Module 1: Lecturer Identification
Module 1 is linked to a database which contains the latest name list of lecturers in UiTM Sarawak and their particulars which will be used for identification purpose. The login page of this module requires the lecturer to provide his user name and password for verification of his identity and authorisation of his access to the application. Both Modules 2 and 3 begin with this same identification process. 3.2 Module 2: Online Preferences Registration
The lecturers will be informed online once the final examinations timetable is uploaded onto the ISS web page. Then, they can register their preferred invigilation time slots via the internal email system. There are five web pages for gathering information from lecturers. Page 1 (Lecturer Identification) which displays the lecturer’s personal particulars will request the lecturer to indicate if all his subjects are based on 100% coursework. A ‘Yes’ ‘Yes’ answer will lead the lecturer to Page Page 2 (Preferred Working Day Selection) to begin the selection process for his preferred time slots. If the lecturer’s answer is ‘No’, he proceeds to Page 3 (Own Subject) where he provides information on all the subjects he teaches, and their respective examination dates and time. After that, the lecturer will be redirected to Page 2 to choose his preferred weekday time slots. Once this is completed, the lecturer will proceed to Page 4 (Preferred Non-Working Day Selection) where he chooses his preferred dates and time for non-working days which will include weekends and any non-working weekdays the examinations are held. The lecturer indicates in a column whether the specific time slot chosen coincides with his own paper. Invigilating on a non-working day is regarded as equivalent to invigilating twice on a working day, so some lecturers prefer to invigilate on non-working day so that they can reduce the number of invigilation duties by one weekday. In a case like this, the lecturer will have to indicate whether he chooses to invigilate on that particular non-working day slot. Page 5 (Comment page) allows the lecturer to include any comments while on Page 6, a summary of all the information entered is displayed. The system allows the lecturer to edit or change any of the entries displayed on the summary page by clicking the ‘edit’ button, and the only data that cannot be changed is the list of subjects he is lecturing. The system will automatically display his entries for each page when he next logs in before the application process is closed. After the specified period for applying for preferred invigilation slots expires, the system will be closed, and the data will be e xtracted from the database of the online system for processing and assigning of invigilation duties. 3.3 Module 3: Invigilation Schedule Retrieval
Once the invigilation schedule is prepared, the relevant information will be uploaded to Module 3 of ISS and then made accessible to all the lecturers. Upon successful login, the system will show a page with instructions for the invigilators from the Campus Campus Director. The lecturer will click the “YES” button at the bottom of the page to indicate that he has read the instructions and understood his responsibility. The system will then lead the lecturer to the next page or else a pop-up message box will appear to instruct him to read the instructions given, and the web page will be redirected to the Responsibilities of Invigilators page. These instructions can be printed using a link provided on the top of the web page. On the following page, the system will retrieve the individual summary of each lecturer’s invigilation duties from the database using the user name in the login page for identification. Here, the lecturer can view and print his individual invigilation schedule. This page also allows the lecturer to submit feedback to the committee regarding his personal schedule. The individual invigilation schedule which is posted in Portable Document Format (PDF) can only be viewed by the lecturer concerned but the final examinations Paper number: [1897541]
CS SR
14 - 15 March 2009
0 8’ 0 9
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
timetable and the master invigilation schedule which contains complete information on the final examinations and invigilation duties can be viewed by everybody. 4. RESULTS AND DISCUSSIONS
ISS is an efficient tool in data collection and information dissemination as it has successfully reduced the mannual involvement and time taken by the scheduling committee in preparing the invigilation schedule. Table 1 compares the online system and the manual way of filling in invigilation preference forms. Table 1: A comparison between ISS and the manual system in filling in invigilation preference forms
Task
Time required for performing the task Online system
Posting of final examination timetable for viewing
Uploading onto webpage
Preparing data in soft format before assigning duties
Downloading from online system
10 mins.
Manual system
Printing and binding timetable
10 mins. Before: Printing of forms for filling in preferences and other information. Preparing boxes for collecting completed forms. After: Sorting, numbering completed forms. Typing and transferring data.
Posting completed invigilation schedule List of invigilators’ duties and responsibilities
60 mins.
4 hrs.
16 hrs.
Uploading onto the webpage
10 mins.
Printing and binding timetable
60 mins.
Uploading onto the webpage
10 mins.
Printing, binding and distributing
8 hrs.
An online survey was conducted amongst the UiTM Sarawak lecturers on their satisfaction level of the effectiveness of ISS as an efficient communicative tool, and whether ISS was user-friendly enough to allow them to enter their invigilation preferences at their own convenience. The responses received were very encouraging judging from the fact that more than 80 per cent of 244 lecturers who participated in this survey agreed that they were very satisfied with what ISS has to offer, and, that is, as a user-friendly and an efficient communicative tool in optimising their invigilation preferences. The analysis of the survey is shown in Figure 2 and Figure 3. Most of the lecturers who indicated their invigilation preferences were often allocated the dates and time they requested for.
Paper number: [1897541]
CS SR
0 8’ 0 9
14 - 15 March 2009
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
Figure 2: The result of the survey question on ‘The online system allows me to enter my invigilation preferences at my own convenience’
Figure 3: The result of the survey question on ‘ISS is an efficient way of communicating with the scheduling committee on my constraints and preferences’ 5. CONCLUSIONS
ISS has proven to be successful in optimising lecturers’ preferences and constraints in examination invigilation besides reducing the manual involvement and time taken to prepare the schedules. Ever since the introduction of ISS in UiTM Sarawak, the system has been constantly upgraded and its automation increased. This innovative online system facilitates communication between UiTM Sarawak lecturers and the ISS committee in optimising the examination invigilation schedules. Paper number: [1897541]
CS SR
0 8’ 0 9
14 - 15 March 2009
C O N F E R E N C E ON S C I E N T I F I C & S O C I A L R E S E A R C H
ACKNOWLEDGEMENT
There are many people who had been instrumental in the completion of this project and the ISS committee regards them in high esteem for being part of the process of creating a more efficient and userfriendlier examination invigilation scheduling system. REFERENCES
Adobe Systems Incorporated. (2008). Dreamweaver (2008). Dreamweaver CS3 resources. http://www.adobe.com/suppo http://www.adobe.com/support/documentation rt/documentation/en/dreamweaver/. /en/dreamweaver/. Online 9 August 2008. Burke, E.K, Bykov, Y. and Petrovic, S. (2000). A multicriteria approach to examination timetabling. In Burke, E. K. and Erben, W. (Eds) Practice and Theory of Automated Timetabling (PATAT 2000, Science. Vol. 2079. SpringerKonstanz, Germany, August, selected paper). Lecture paper). Lecture Notes in Computer Science. Verlag, Berlin Heidelberg New York. 118-131. Burke, E.K, Elliman, D., Ford, P. and Weare, R. (1996). Examination (1996). Examination timetabling timetabling in British universities a st survey. survey. 1 International Conference on the Practice and Theory of Automated Timetabling (ICPTAT ’95, Napier University, Edinburgh, UK, 30 th Aug – 1st Sept 1995). Dahal, K. P. Aldridge, C. J., McDonald, J. R. (1999). Generator maintenance scheduling using a genetic algorithm with a fuzzy evaluation e valuation function. Fuzzy function. Fuzzy Sets and System, 102: 21-29. Hishammuddin Asmuni, Burke, E. K. and Garibaldi, J. M. (2004). Fuzzy multiple ordering criteria for examination timetabling. PATAT (2004). Practice (2004). Practice and Theory of Automated Timetabling . Pittsburgh, PA USA. August 18-20, 2004. Junginger, W. (1986). Timetabling in Germany – A survey. Interfaces survey. Interfaces,, vol. 16. no. 4, pp. 66-74. In Burke, E., Elliman, D., Ford, P. and Weare, R. (1996). Examination timetabling in British universities a survey. survey. 1st International Conference on the Practice and Theory of Automated Timetabling (ICPTAT ’95, Napier University, Edinburgh, UK, 30 th Aug – 1st Sept 1995). Li, J. and Kwan, R. S. K. (2003). A fuzzy genetic algorithm for driver scheduling. European Journal of Operational Research, 147: 334-344. Ong, M. L., Liew, L. H. and Sim, J. (2008). Examination Invigilation System – An Online Approach. Proceedings of UiTM Sarawak Conferenmce 2008.UiTM Malaysia:UPENA. Malaysia:UPENA. Ong, M. L., Liew, L. H. and Sim, J. (2006). Examination Invigilation Scheduling System In Optimising Lecturers’ Preference. UiTM Sarawak: Unit of Research, Development and Commercialization (URDC), Sarawak: Universiti Teknologi Mara. Raj, P. A. and Kumar, D. N. (1999). Ranking alternatives with fuzzy weights using maximizing set and Systems . 105: 365-375. minimizing set. Fuzzy set. Fuzzy Sets and Systems. Teodorovic, D. and Lucic, P. (1998). A fuzzy set theory approach to the aircrew rostering problem. Fuzzy problem. Fuzzy Sets and Systems. Systems. 95: 261-271.
Paper number: [1897541]
