Introduction
Timetable scheduling in universities has experienced numerous difficulties in the past decades, with subject allocation, assigning time slots, and classrooms posing the greatest challenge. Such challenges have made programmers use concepts like optimization to enhance the scheduling process. Due to different reasons, optimization has become a required field for solving several problems, their types, and their degrees of complexity.
In 2017, Soyemi et al. stated that there are difficulties relating to UCTS, where individuals seek to maximize the satisfaction of teachers and students and minimize the empty time. Different studies outline the manner in which universities utilize optimization methods and their impact on timetable scheduling activities. This chapter will systematically examine various studies outlining the challenges of timetable scheduling and the provision of literature on how generic algorithms are used to solve the problem.
Timetable Scheduling Problem in Universities
The increase in the number of students joining universities around the globe has forced institutions to obtain more personnel and resources to meet educational demands. The pressure to develop courses and lessons that meet the needs of the students has made the call for reforms inevitable. In 2011, Dong et al. stated that it has become essential that the teacher, curriculum, and student be assigned an appropriate time section in a formal classroom. The terminologies used include; the timetable problem (TTP), simulated annealing algorithm, and genetic algorithm.
The results of the article state that the main advantages of using algorithms as the main technology is that it enables the programmers to eliminate early convergence and variation problems are the main challenges linked to the approach. Despite such problems, there is the possibility of improvement due to its genetic nature, making it the most appropriate tool for timetable scheduling. Additionally, in [2011], Dong et al. stated that it is clear that the computerization of university systems led to the development of automated course scheduling algorithms, which eased the problem. The main disadvantage is that automated course scheduling algorithms are linked to an exponential increase in computing time, integration time, and space.
Research shows that automated course scheduling has been utilized in colleges and universities since the 70s. In 2017, Soyemi et al. stated timetabling problem is perceived to enhance over the years as courses and student enrolment increased. Computer application is the main technology used in the article, and the authors examine how the iterative adaptive heuristic probabilistic algorithm has enhanced the process of timetable scheduling. In 2017, Soyemi et al. stated that the approach has a global optimization strategy that generates codes under optimal conditions for arranging and then selects the appropriate function to measure their degree of fitness. The main advantage of the technology is that it satisfies the various constraints making it effective in carrying out heredity operations as an exchange, choice, or variation to the individual. The technology is also quicker than most other algorithms, fault-tolerant, simple and strong. The disadvantage is that the genetic algorithm (GA) does not directly act on the variables but on individuals.
Over the years, different mechanisms have been implemented to optimize timetabling processes. In 2017, Soyemi et al. stated that the main advantage of the technology is that the algorithm mimics the annealing process to attain a global optimization solution. In other colleges, a backtracking algorithm has been used to provide a combination of solutions for a large but limited number, enhance the process and reduce the number of conflicts experienced. The main challenge with the backtracking algorithm is time complexity making it not perfect to be used alone but with other algorithms. Additionally, in 2019, Abayomi-Alli et al. stated that approaches like particle swarm optimization have been highly used due to their population-based nature, which adopts the social behavior of fish or bird flocking schooling. In the article, the authors use particle swarm optimization as the main technology used to solve the timetable scheduling problem. The article results show that the concept comprises acceleration or changing the velocity of each particle or random term towards gBest and lBest locations.
Other studies show that course scheduling problem encompasses factors like time, class, grade, teacher, and the curriculum. In 2019, George & Vijayan and Abayomi-Alli et al. stated that such elements need to be examined to ensure that the institution enhances its productivity and meets its objectives within the limited time and with the available resources. The software used to attain automatic course scheduling is expected to achieve an optimal arrangement of courses to ensure no conflict is experienced. Educational timetabling challenges are perceived to take three primary forms: school, examination, and course timetabling.
Constraints in Course Scheduling in Universities
Constraints form the most critical part of efficient course scheduling in universities. Timetable constraints refer to the rules or conditions that are used to manage available resources. In 2015, Chauhan et al.. Such constraints enabled the avoidance of time clashes and enhanced resource workloads’ suitability. Timetable constraints are divided into hard and soft constraints based on their impacts on the timetable’s effectiveness. Hard constraints are perceived to have a free error on timetables that are experienced as free time classes and enable avoidance of resource overload. Soft constraints develop selective rules that are applied in the timetable scheduling to enhance the efficacy of courses and reduce the need for part-time lecturers. Soft constraints are also used to provide an off day for the lecturers every week.
Apart from hard constraints, a variety of soft limitations are also used to enhance the effectiveness of schedules. In 2015, Chauhan et al. stated that such constraints are applied to honors, and general courses need to be organized in time slots that are not overlapping. In 2016, Pandey argued that other soft constraints entail spreading the available lecturers within the minimum days of the course, giving consecutive lecturers in a given curriculum, and opening courses to a minimum number of registered students. In addition, in 2019, George & Vijayan and Abayomi-Alli et al. stated that faculty courses must be allotted first slots before the departmental courses; students should be given the first preference for having an exam a day, and breaks should be allotted slots before other courses. Due to this explanation, search GA enables the development of schedules that meet the various constraints, making the timetable scheduling highly effective and covering all the needs of the students and the lecturers.
Research on constraints shows that various constraints are used to enhance the effectiveness of timetable planning. In 2019, George & Vijayan, several hard constraints influenced the efficacy of university schedules. The study’s example is that courses with common students should not be allotted time within the same day if the total number of sessions is 8 hours. Most hard constraints in developing university scheduling programs include not assigning more than one instructor to a given room in the same time slot and not assigning one different lecturer courses at the same time. Further, they involved not assigning the same class to two courses simultaneously. Other hard constraints include developing time plans that meet the university calendar, not developing a course that exceeds the maximum number of students, and developing unique courses.
Timetable Investigations
Timetabling problems is among the various aspects taught in machine learning. The use of algorithms enables the developers to attain optimization in assigning lectures courses in the available time within the term. In 2021, Peng et al. stated that an optimization problem is used in finding the set of parameters allowing the timetable to obtain the best result. The article uses mathematical optimization to investigate the timetable scheduling problem. Results show that the concept involves translating the optimization problem, after a problem analysis, into an equivalent mathematical problem. This is the most delicate step of the resolution process because the formulation of a problem is never unique; in particular, the definition of the functions characterizes the system’s performance.
It is presumed that public universities take days to schedule all the classes students should take manually. In 2021, Arratia-Martinez et al., such scheduling procedures consider the number of lecturers available and classrooms. In such cases, it is advised that public institutions take into consideration the automation of the timetable scheduling process to save time and meet the real needs of the institutions. Such mechanisms are used to solve the complex problems that arise in the scheduling process, like the availability of essential resources like classrooms, students, and lecturers.
The research on the university timetable problem encompasses two broad educational aspects, which entail the necessary teaching tasks fortitude and the optimization of the ordinary teaching time. In 2021, Kakkar et al. stated that using GA provides detailed information on various steps to optimize the teaching time and enhance the fitness function variations, genetic crossover, and the initial population generation. Further, Kakkar et al. argued that the algorithm ensures that it takes care of all constraints and arranges an optimal schedule with available lecturers, many classes, and resourceful courses. A highly constrained combinational problem like the timetable can be solved through aspects like evolutionary methods like genetic algorithms. Such methods are perceived to play a significant role in ensuring that the timetables have the minimum number of conflicts possible.
Generic Algorithm Investigations
A study by George and Vijayan in 2019 shows that GA is among the commonly used optimization methods for timetable scheduling. The study indicates that the method is used together with other methods like particle swan optimization, fuzzy logic, and ant colony system. In 2019, George & Vijayan stated that the main advantage of the article is that it outlines how GA has been effective in developing classroom schedules considering resources and parameters like priority values for teachers, subjects, and hours. The authors mention that GA makes scheduling easier and takes a shorter time. The study’s main limitation is that it uses a computerized experiment to examine the effectiveness of the optimization method.
Over the years, several methods for solving optimization problems have been proposed. In 2020, Alghamdi et al. claimed that these methods are grouped into two classes: the class of exact methods and the approximate method. Exact methods obtain optimal solutions and guarantee their optimality; however approximate methods sacrifice the solution’s optimality by obtaining a more reasonable time. On the other hand, approximate methods generally provide an acceptable solution in a reasonable time that balances time and acceptance of the solution. Additionally, Alghamdi et al. argued that these two classes of algorithms show a heuristic and metaheuristic subsection. Heuristics make it possible to approach instances of large problems by providing satisfactory solutions within a reasonable time.
Other studies also examine the various constraints that timetable scheduling and how GA enables solving such issues. In 2020, Alghamdi et al. outlined timetabling restrictions as either essential or non-essential, where the essential limitations are perceived as factors that make a timetable unworkable if not satisfied. The studies argue that GA borrows significantly from natural sciences, allowing it to observe challenges from the lens of natural phenomena. In 2015, Chauhan et al. claimed that GA enables it to develop computational mechanisms resembling mutation and natural selection easily. Such mechanisms make it easy for developers to enhance the proper functioning of the algorithm. The mutation process enables the avoidance of aspects like stagnation within a population which enhances the formulation of an optimization model.
This process involves the use of mathematical algorithms to solve the challenges linked to the timetable scheduling process. It consists of precisely defining the genetic algorithm and the objective to be achieved and can be of two types: a cost to be minimized or a performance to be maximized. In 2021, Peng et al. claimed that in terms of performance, the scheduling process should maximize the returns and transmittance to ensure minimum conflicts are experienced in the assignment of lessons. Studies on the effectiveness of Genetic algorithms show that it take various forms: geometric dimensions, resource properties, and maintaining structural choices. Such optimization mathematics can be quantitative or qualitative, continuous or discrete. The algorithm also describes all the hard constraints related to the problem posed.
Benefits of Genetic Algorithms
A genetic algorithm is a combinatorial optimization algorithm used to solve the timetable scheduling problem. The problem can determine what the automatic generation of schedules can achieve by analyzing the results of the various published studies. In 2015, Chauhan et al. stated that comparing the algorithm-generated schedules with those in real-world learning institutions, some studies have found that the algorithm-generated schedules are of higher quality since they incorporate some evaluation functions. The process utilizes various techniques like neural networks to solve timetabling problems since there is no specific efficient way to solve the issue.
Generic algorithms are beneficial since they can be easily parallelized in terms of either data or control. Data parallelism entails the execution of a similar procedure for multiple data sets at the same time. On the other hand, in 2020, Alghamdi et al. stated that control parallelism is attained when computer codes can be used to distribute tasks across different processors. Another benefit of genetic algorithms is that it is a great global optimization tool normally used to eradicate the problem of complicities when applied correctly. The probabilistic nature of the model makes it effective in providing a solution through a genetic presentation of chromosomes that can be easily interpreted and analyzed.
Conclusion
The chapter outlines how universities experience challenges in the timetable scheduling process. It outlines how the complexity of developing inclusive timetables and the various mechanisms that provide solutions to such challenges have changed the timetable scheduling process. They have changed the process’s computerization with time and how programmers have used mathematical concepts to optimize the available resources and solve the complex problem of timetable scheduling. Lastly, the various advantages linked to genetic algorithms are discussed to enhance the understanding of how the processes are beneficial.
Table of Article Comparisons
References
Abayomi-Alli, O., Abayomi-Alli, A., Misra, S., Damasevicius, R., & Maskeliunas, R. (2019). Automatic examination timetable scheduling using particle swarm optimization and local search algorithm.Data, Engineering and Applications, 119-130.
Alghamdi, H., Alsubait, T., Alhakami, H., & Baz, A. (2020). A review of optimization algorithms for university timetable scheduling.Engineering, Technology & Applied Science Research, 10(6), 6410-6417.
Arratia-Martinez, N., Avila-Torres, P., & Trujillo-Reyes, J. (2021). Solving a university course timetabling problem-based on AACSB policies. Mathematics, 9(19), 2500.
Chauhan, P. M., Parmar, K. B., & Mendapara, M. B. (2015). Solving timetable scheduling problem by novel chromosome representation using genetic algorithm. In 2015 International Conference on Circuits, Power and Computing Technologies [ICCPCT-2015] (pp. 1-6). IEEE.
Dong, Y., Qi, B., Gao, W., Wang, Q., & Wang, L. (2011). Algorithms of timetable problem application and performance evaluation in course scheduling system. In 2011 2nd International Conference on Artificial Intelligence, Management Science and Electronic Commerce (AIMSEC) (pp. 3702-3705). IEEE.
George & Vijayan. (2019). Multi-objective timetable scheduling using genetic algorithm. International Journal of Networks and Systems, 8(3), 92-94. Web.
Kakkar, M., Singla, J., Garg, N., Gupta, G., Srivastava, P., & Kumar, A. (2021). Class schedule generation using evolutionary algorithms. Journal of Physics: Conference Series, 1950(1), 12067. Web.
Pandey, H. M. (2016). Solving lecture time tabling problem using GA. In 2016 6th International Conference-Cloud System and Big Data Engineering (Confluence) (pp. 45-50). IEEE.
Peng, W., Mu, J., Chen, L., & Lin, J. (2021). A novel non-dominated sorting genetic algorithm for solving the triple objective project scheduling problem.Memetic Computing, 13(2), 271-284.
Soyemi, J., Akinode, J., Appiah Osei, M., & Oloruntoba, S. (2017). Automated lecture time- tabling system for tertiary institutions. International Journal of Applied Information Systems, 12(5), 20-27.