Computational Social Choice for Social Good: Altruism in Matching Problems with Preferences

The thesis provides an introduction to Computational Social Choice, that is an interdisciplinary field at the interface of social choice theory and computer science, tackling topics such as fairness, strategic manipulation, and altruism, highlighting their relevance in actual modern multi-agent systems. It considers in particular matching problems with preferences, that are problems that occur in many applications. The task of these problems is to find a matching of the participants that is in some sense optimal with respect to these preferences. The thesis explores enhancements to the matching framework with a focus on the integration of the altruistic behavior of the agents in this framework. A novel contribution of the thesis is the adaptation of the Gale-Shapley algorithm, a classical algorithm for solving matching problems with preferences, in order to accommodate altruistic agents. The thesis also provides an experimental evaluation to evaluate how altruism influences the quality of the resulting matching, demonstrating in general an overall improvement in satisfaction of agents.