Abstract:

Every year, students in Higher Education face the challenge of carrying out a long-term project that encompasses a wide breadth of the skills and competences developed during their respective degrees. These projects usually come in the form of undergraduate, master, or PhD. dissertations. While there is a wide range of strategies, tools, and methodologies that have been studied with the purpose of conducting good in-class and module/course experiences, there is, comparatively, very little effort on the matter of conducting the experience and guidance of students with respect to their individual dissertations. Yet, we believe that this is an important event in every student's academic life, as it usually entails their degrees' completion and an opportunity to exhibit their accomplishments.

Generally, the responsibility of assigning students to academic supervisors lies on the shoulders of decision makers in academic management (i.e., department heads, coordinators, etc.). Albeit it may initially seem simple, decision makers face an intricate problem influenced by a variety of criteria such as the number of students to allocate, the number of supervisors from where to choose, both students' and supervisors' preferences on research topics, work load constraints by academic staff, and even the department's social climate. As a result, the decision on how students should be allocated to supervisors is arduous, time consuming, and, quite often, stressful.

This decision has almost limitless possibilities and, making an optimal decision on the matter, may be unfeasible for humans. Despite this problem, Artificial Intelligence has proved to be a helpful tool in supporting decision making in a variety of complex domains such as emergency response, analytics, health, sports, or even education. Its ability to manage large decision problems (i.e., involving many variables), provide time effective responses, and learn from experience has put artificial intelligence to the forefront of today's ICT solutions. We believe that the problem of assigning students to supervisors is a complex one, and it would benefit from the support of Artificial Intelligence tools.

In this paper, we present an artificial intelligence tool that aims to support members in academic management in the decision of allocating students to supervisors for their academic dissertations. On the one hand, the tool takes into consideration the preferences of both students' and supervisors' on research topics to ensure that students are advised by supervisors with experience on their desired research topic, and that supervisors feel comfortable advising and guiding students throughout their dissertations. On the other hand, the tool also takes into consideration the different workload levels of individual supervisors. That is, the maximum number of students that each supervisor may be able to supervise. In order to foster a good work climate, the tool also ensures that all supervisors have a similar workload in the final allocation. We have implemented some pilot experiences, with positive perspectives, to test its feasibility as a decision support tool for members in academic management.

Keywords:

Apps for education, New projects and innovation, Academic management, Artificial Intelligence.