Hybridization and artificial intelligence in optimizing university examination timetabling problem: A systematic review

University Examination Timetabling Problem is the most important combinational problem to develop a conflict‐free timetable to execute all of the exams in and with the limited timeslots and other resources for universities, colleges or schools. It is also an important Nondeterministic Polynomial Time (NP)‐hard problem that has no deterministic solution, so it is an important study to develop an optimised solution to satisfy all of the hard and soft constraints required to create a timetable that would be flexible for all stakeholders such as students, teachers, invigilators, as well as management. Several heuristics, meta‐heuristics and mixed integer programming approaches have been performed in the past to develop a solution for it, but use of hybrid techniques and hyper‐heuristics through implementing Artificial Intelligence (AI) improve the efficiency and performance in solving the problem using the single or multiple objective function. In this systematic review, we are going to learn the effectiveness of these heuristic and meta‐heuristics techniques in solving the examination problem as well as how these can be merged to develop a more effective and efficient hybrid solution. The impact of AI is also reviewed in developing the solution for the examination timetabling problem. While analysing papers included in this SLR, various research gaps are identified including the level of hybridisation between population‐based and local‐search techniques, designing an effective algorithm for an initial solution, providing a balanced solution for multiple stakeholders, and developing a phase‐wise and partial solution to improve the effectiveness of a final algorithm for the university examination timetabling problem. It provides insight into the various challenges and opportunities as well as future directions to develop or to perform research in providing a solution for it. It is identified that the use of machine‐learning, reinforcement learning, and deep‐learning with more advanced AI techniques can optimise and to improve the performance of a solution for the university examination timetabling problem. Context and implications: Rationale of this study: This literature review was conducted to present an efficient and optimised solution for the university examination timetabling problem (UETP) by implementing hybridisation and artificial intelligence (AI) techniques. The objective is to maximise satisfaction for students, teachers, invigilators and management to ensure examination activity in universities without any resource conflict. Why the new findings matter: The review highlights the positive impact of hybridisation and AI techniques including machine‐learning, deep‐learning and reinforcement learning by exploring the new dimensions to develop or enhance an algorithm that provides an optimised solution for the university examination timetabling problem with greater efficiency and accuracy. Implications for researchers and educational institutions: The Implication of this study is significant for universities or educational institutes in providing an efficient and optimised technique to conduct examination activities. It provides a comprehensive guide to minimise all of the conflicts in resource allocation during examination activities, ensuring a satisfactory and optimised solution for all stakeholders including students, teachers, invigilators and management to a greater extent. This research offers an opportunity to researchers to explore the application of hybridisation and AI techniques, including machine‐learning, deep‐learning and reinforcement learning in developing and enhancing an efficient and optimised algorithm for the university examination timetabling problem.