From Integer Programming to Machine Learning: A Technical Review on Solving University Timetabling Problems
and Zongwen Fan 5,‡
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThis review summarizes the findings from a technical assessment of university timetabling problems solved using integer programming algorithms. To improve the article's quality, the following recommendations are made:
1. The article's readability, particularly in the abstract, needs attention. The innovation and contribution should be presented more concisely, as the current version is overly verbose. Listing the contributions in bullet points would increase clarity. Additionally, the conclusion should be more succinctly summarized.
2. The literature review is somewhat outdated. It is recommended that the author explore more recent works from the past three years. In particular, recent research on “joint service deployment and task scheduling” and “collective edge intelligence sharing” should be included in the discussion.
3. Existing research should not merely be listed but summarized, with personal insights and opinions included.
4. The fourth section needs more detailed elaboration instead of general statements.
5. Finally, a discussion on the challenges of future openness should be included.
Author Response
1. The article's readability, particularly in the abstract, needs attention. The innovation and contribution should be presented more concisely, as the current version is overly verbose. Listing the contributions in bullet points would increase clarity. Additionally, the conclusion should be more succinctly summarized.
Response: Thanks for this comment. We have made extensive changes to the abstract, introduction and conclusion and also reviewed all other parts of the paper. The changes are highlighted in red.
2. The literature review is somewhat outdated. It is recommended that the author explore more recent works from the past three years. In particular, recent research on “joint service deployment and task scheduling” and “collective edge intelligence sharing” should be included in the discussion.
Response: Thanks for this comment. The literature review is extended with machine learning based contributions in the area, six new references are added to the paper.
3. Existing research should not merely be listed but summarized, with personal insights and opinions included.
Response: Thanks for this comment. We have added authors’ insight in section 2 for all reviews, changes highlighted in red.
4. The fourth section needs more detailed elaboration instead of general statements.
Response: Thanks for this comment. We have added elaborations in section 2 for all reviewed papers, changes highlighted in red. A summary paragraph is added to explain RQ4. Discussion on research questions is extended with RQ5 andRQ6.
5. Finally, a discussion on the challenges of future openness should be included.
Response: Thanks for the comment. Future work is updated in the Conclusion section.
Reviewer 2 Report
Comments and Suggestions for AuthorsThe paper provides a technical review on solving university timetabling problems (UTP) using integer programming algorithms. It identifies key challenges involved in scheduling courses, rooms, and other academic activities while minimizing conflicts and optimizing resource usage. The authors focus on Integer Linear Programming (ILP) techniques, discussing how these models help manage constraints such as class availability, instructor schedules, and room capacities. The study also compares existing solvers, like CPLEX, used in timetabling research, and highlights gaps in the literature. It concludes by suggesting future research directions, emphasizing the potential of machine learning and agent-based modeling for developing more adaptive solutions to UTP. The following major concerns should be addressed to improve the quality of the manuscript.
· The paper outlines four key research questions, but some overlap between them (e.g., RQ2 and RQ3) could confuse readers. Consider rephrasing or consolidating them to ensure
· Although the paper provides a good overview of previous research, some recent studies on metaheuristics and AI applications in university timetabling seem missing. Incorporating more contemporary comparisons would strengthen the review's relevance and completeness.
· The use of only the Scopus database may limit the breadth of the literature. Including other databases (e.g., IEEE Xplore or Google Scholar) could improve comprehensiveness and ensure no relevant work is overlooked.
· There are minor formatting inconsistencies in the text, such as capitalization errors in section headers (e.g., "Classroom timetabling" vs. "course Timetabling"). A careful proofreading pass is needed to maintain consistency.
· The paper suggests future work using machine learning and agent-based modeling but provides limited detail on how these methods could be integrated with integer programming. Including more specific examples or frameworks would enhance the practical value of these suggestions.
· The section comparing solvers focuses primarily on frequency of use (e.g., CPLEX used in 47 studies). It would be beneficial to expand on performance metrics (e.g., solution quality, time efficiency) to provide a more comprehensive comparison of solver effectiveness.
Comments on the Quality of English Language
Moderate editing is required.
Author Response
1. The paper outlines four key research questions, but some overlap between them (e.g., RQ2 and RQ3) could confuse readers. Consider rephrasing or consolidating them to ensure
Response: The RQ3 is changed from what solvers to what tools to differentiate RQ2 and RQ3. Two new research questions are added as well.
2. Although the paper provides a good overview of previous research, some recent studies on metaheuristics and AI applications in university timetabling seem missing. Incorporating more contemporary comparisons would strengthen the review's relevance and completeness.
Response: Thanks for the valuable feedback. Machine learning solutions in university timetabling are added to the literature review in section 4. It makes the entire review more interesting.
2.a. The use of only the Scopus database may limit the breadth of the literature. Including other databases (e.g., IEEE Xplore or Google Scholar) could improve comprehensiveness and ensure no relevant work is overlooked.
Response: Thanks for this comment. The comparison among Web of Science, Google Scholar, and Scopus are added to the methodology section. The coverage of Scopus is also explained in the methodology section.
2.b. There are minor formatting inconsistencies in the text, such as capitalization errors in section headers (e.g., "Classroom timetabling" vs. "course Timetabling"). A careful proofreading pass is needed to maintain consistency.
Response: Thanks for this comment. We have reviewed the article and such inconsistencies are removed.
2.c. The paper suggests future work using machine learning and agent-based modeling but provides limited detail on how these methods could be integrated with integer programming. Including more specific examples or frameworks would enhance the practical value of these suggestions.
Response: Thanks a lot for the valuable comments. The literature review is extended to the machine learning. The conclusion is also updated accordingly.
2.d. The section comparing solvers focuses primarily on frequency of use (e.g., CPLEX used in 47 studies). It would be beneficial to expand on performance metrics (e.g., solution quality, time efficiency) to provide a more comprehensive comparison of solver effectiveness.
Response: Thanks for this comment. Research question 3 is updated and now performance comparison is included in the comparison of the tools.
Reviewer 3 Report
Comments and Suggestions for AuthorsPlease see the attached report
Comments for author File: 
Comments.pdf
Author Response
1. Provide a more detailed explanation of how integer programming differs from other optimization techniques, specifically in the context of university timetabling.
Response: Thanks for this comment. In section 4 we have added detailed explanation how integer programming differs from other optimization techniques, changes are highlighted in red.
2. In this review article authors only use Scopus. Are there any potential limitations or weaknesses in using only Scopus as the bibliographic source for this review. Would the inclusion of other databases strengthen the findings?
Response: Thanks for this comment. Explanation is added to the first paragraph of the methodology section.
3. The article mentions several commercial solvers (e.g., CPLEX, Gurobi). Provide a more detailed comparative analysis of these solvers, including performance metrics.
Response: Thanks for this comment. Research question 3 is updated and now performance comparison is included in the comparison of the tools.
4. The author needs to include a section or subsection discussing the scalability of integer programming models when dealing with larger, more complex timetable scenarios? How can these models be improved for such cases?
Response: Thanks for this comment. In section 4.3, we have included explanation on scalability issues faced by integer programming approaches and also on some potential solutions of those issues, changes highlighted in red.
5. The review focuses on integer programming algorithms. Include a brief overview of alternative methods, such as metaheuristic approaches, and explain why these might be less favorable.
Response: Thanks for this comment. In section 4.3 comparison of integer programming with other approaches is provided, highlighted in red.
6. The author may also include a paragraph detailing the challenges of integrating modern technologies like machine learning with integer programming in university timetable.
Response: Thank you very much for this valuable advice. Literature review is extended to include machine learning based approaches for solving timetabling problems.
7. This review identifies different types of timetable problems (course, classroom, and exam). How did the authors ensure they covered all major categories comprehensively.
Response: Thanks for this comment. These are the major subcategories identified in the literature and we have mapped our observations to those categories, references provided in the text.
8. In the methodology section, what specific criteria were used to determine whether an article was relevant or irrelevant. Add some more explanation.
Response: Thanks for this comment. We have added selection criteria in methodology section to elaborate on our inclusion and exclusion criteria.
9. In this review, the authors highlight that CPLEX is the most used solver. Is there any evidence explaining why CPLEX is preferred over other solvers, such as Gurobi or Lingo.
Response: Thanks for this comment. Performance comparison of solvers is added to RQ4 to elaborate on the strengths and weaknesses of these tools.
10. This review mentions a high implementation rate for integer programming solutions. How do the authors ensure that this implementation rate is consistent across different regions or only within certain academic contexts.
Response: Thanks for this comment. RQ3 section is updated to explain the implementation rate of these solutions.
11. How did the author validate the accuracy and relevance of the datasets mentioned? Could they provide more information on how these datasets were benchmarked?
Response: Thanks for this comment. More information on the datasets is included for benchmarked comparisons.
12. Authors may look for some punctuation, typos and editing issues.
Response: Thanks for this comment. We have reviewed the paper and corrected the typos and editing issues.
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsThis is a highly unsatisfactory revision of the manuscript. Many issues remain unresolved. The authors have failed to address the reviewers' comments point by point. Moreover, the references contain garbled characters.
Comments on the Quality of English LanguageThis is a highly unsatisfactory revision of the manuscript. Many issues remain unresolved. The authors have failed to address the reviewers' comments point by point. Moreover, the references contain garbled characters.
Author Response
All changes are applied in Round 1. Please check the latest manuscript.
Reviewer 2 Report
Comments and Suggestions for AuthorsAuthors have addressed all my concerns, it can be accepted for publication.
Comments on the Quality of English LanguageMinor editing is required.
Author Response
Thank you for the comments.
Reviewer 3 Report
Comments and Suggestions for AuthorsAll the suggested changes have been incorporated in the revised version
Author Response
Thank you for the comments.
