Development of a Data-Driven Class and Thesis/Capstone Defense Scheduling Decision Support System for Higher Education
This study presents a web-based decision support system that integrates class and thesis/capstone defense scheduling through a constraint-aware, rule-based workflow, demonstrating high user satisfaction and operational efficiency in deployments across two higher education institutions.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Every university faces a recurring puzzle that feels less like a math problem and more like a high-stakes game of musical chairs. It is the task of fitting people, places, and time into a single, coherent schedule. Administrators must assign professors to classrooms, match students to courses, and ensure that no two events try to occupy the same room or the same professor at the same moment. For decades, many schools have solved this by hand, using spreadsheets, whiteboards, and sheer persistence. This manual approach is slow, prone to mistakes, and often leaves administrators blind to the bigger picture, such as whether a professor is secretly overworked or if a specific type of classroom is being used to its absolute limit. While computers have long been able to generate timetables, they have rarely been able to handle two critical types of events at once: the daily rhythm of regular classes and the special, high-stakes moments of thesis or capstone defenses. These two worlds usually operate in separate silos, forcing humans to cross-reference lists to avoid double-booking a professor who is teaching a lecture in the morning and judging a student defense in the afternoon.
A team of researchers at Sorsogon State University in the Philippines has built a new kind of digital tool designed to bring these two worlds together. They created a web-based system that acts as a central command center for academic scheduling. Instead of trying to invent a complex new mathematical formula to solve the puzzle, they focused on building a smart, rule-based workflow that understands the specific constraints of a real university. The system treats faculty availability, room capacity, and student needs as a single, connected dataset. When an administrator asks the system to generate a schedule, it checks every rule at once: it ensures a professor is not in two places at once, that a room is big enough for the class, and that a thesis panel is available without conflicting with their teaching duties. The researchers tested this system in two different colleges, one public and one private, to see if it could handle the messy reality of different school cultures and program structures.
The system works by gathering all the necessary information—profiles of every professor, details about every classroom, and the requirements for every course and defense group. It then uses a step-by-step process to build a schedule. It picks an event, finds a time and place that fits all the hard rules, and assigns it. If a conflict arises, such as a professor being needed for two things at once, the system flags it for a human to review rather than forcing a bad solution. This approach is distinct from the complex optimization algorithms often found in academic literature, which try to find the single "perfect" mathematical answer. Instead, this tool prioritizes practicality and transparency, offering a clear view of where conflicts exist and allowing administrators to make informed decisions. The researchers found that by integrating the scheduling of regular classes with the scheduling of thesis defenses, they could spot problems that would have been invisible if the two were handled separately. For instance, the system successfully identified a situation where a professor was assigned to twelve thesis panels in a single week while also carrying a heavy teaching load, a conflict that a standard class scheduler would have missed.
The project was developed using a modern, flexible approach to software creation, building the system in stages and refining it based on feedback from the people who would actually use it. The team started with a database design that proved too complicated for the strict rules of academic scheduling, so they switched to a different type of data storage that enforces strict connections between professors, rooms, and courses. This change prevented errors like assigning a class to a professor who no longer exists in the system or booking a room that has been closed. The final product is a clean, web-based interface where administrators can see the entire university schedule at a glance, color-coded to show available slots, potential conflicts, and room usage. The system was deployed in December 2025 at the College of Information and Communications Technology and in January 2026 at a private college of criminology. In the first location, the system managed over 1,200 class meeting blocks and 142 thesis defenses, resolving hundreds of potential conflicts before the final schedule was printed. It achieved an average room usage rate of 87 percent, indicating that the system was effectively filling available space without overloading it.
To understand how well the system worked, the researchers asked a group of thirteen people—three technical experts and ten university staff members who actually manage schedules—to evaluate it. They used a standard set of questions to measure how useful, fast, easy to use, and reliable the system felt. The results were overwhelmingly positive. The users rated the system very highly for its ability to generate conflict-free schedules, assign panels correctly, and follow the complex rules of the university. They found the interface intuitive and the system responsive, even when handling large amounts of data. While the researchers could not provide some of the deep technical performance numbers usually expected in computer science papers because the raw data from initial testing was not saved, the operational records showed the system was stable. It recovered quickly from a few technical glitches during testing, and security checks confirmed that the system was safe from common online threats. The most significant finding was not that the system solved a mathematical equation faster than any other, but that it successfully unified two separate administrative tasks into one smooth process.
The study highlights a practical shift in how educational technology can be applied. Rather than chasing the most complex algorithm, the researchers demonstrated that a well-designed, rule-based system that integrates different types of data can solve real-world administrative headaches. The system proved that it is possible to build a tool that understands the nuance of faculty workloads and the specific needs of thesis defenses, offering administrators a single place to see the whole picture. The researchers noted that the system's success relied on its ability to adapt to different institutions without needing to be completely rebuilt for each one. However, they also acknowledged that the study was limited to two schools in one province and that the evaluation relied on the opinions of a small group of users. They emphasized that while the system worked well in these settings, other universities would need to test it against their own specific rules and structures. The work serves as a reminder that in the complex world of higher education administration, sometimes the most effective solution is not a magic bullet, but a clear, integrated view of the resources at hand.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.