A Demonstrator to “Expose” Agent-based Argumentation

The potential educational power of digital games which equip learners with motivation and help them focus has well been studied. However, few applications taking advantage of such findings exist, and even fewer are put into practice. In this thesis we investigate the possibility of building game-like applications in order to demonstrate non-trivial AI concepts to non-expert humans. We do so by first implementing Argudem, a game-like demonstrator of agent-based argumentation in the context of automated planning. Then we invite users from a wide range of educational backgrounds to use Argudem so that we can monitor and evaluate their performance. The results of the experimental data analysis suggest that popular demonstrators can be implemented and serve as valuable tools for educational and demonstration purposes. Argudem can serve as a showcase for such applications and motivate initiatives for implementations of similar objective.