Abstract:

The fair division and collective well-fare have been subjects of discussion since the beginning of the civil society. The first notes of formalizing the definition of distributive fairness was given by Aristotle - two persons having the same characteristics regarding a given allocation problem, should be treated equally.

This research is about discrete subtasks fair division within student projects, a problem that cannot be solved by a standard cake-cutting approach, but nevertheless must fulfill the general features of fair division as envy-freeness, equity and Pareto optimality. Opposite of the general cake-cutting strategy where the agents involved aim to satisfy the appetite by maximizing the portion assigned, in our case we deal with agents (students) that have the affinity either to minimize the work or to delegate it to the other members of the team.

The project tasks can be divided in finite number of homogeneous subtasks. However, the problem gets more complicated by the fact that the teams are comprised of students with different course performance background. Allowing the students to choose subtasks by their own preferences is very likely to end up with insincere choice that is not consistent with their true preferences (skills).

We are investigating the most appropriate allocation of subtasks within student’s projects which are compulsory part of the Intelligent Systems course taught at the Faculty of Computer Science and Engineering in Skopje, North Macedonia. The past experience has shown that the students tend to choose the subtasks they find to be the easiest. To avoid the "easiest route" approach, we propose a fair division strategy for solving this problem. The main goal is to guarantee the success and high quality of the student projects. To achieve the aims, our methodology is focused on objective determination of the students' preferences for each project subtask and map them in the most appropriate one.

The results have shown that our approach ensures efficiency and envy-freeness, whereas the equitability is partially proved due to the subjective student’s evaluation of the personal preferences as a result of the biasness to the teachers valuation of the subtasks.

Keywords:

Fair division, tasks, student projects, algorithm.