Verifying the Compliance of Argumentation Principles with Human Reasoning

Over the years, the theory of computational argumentation has drawn the attention of scholars in many fields. It is also used in many AI technologies to justify decisions to end-users. These argumentation-based applications mainly rely on specific techniques that satisfy multiple argumentation principles, which in turn ensure their rational functioning.

However, it is still unknown whether these argumentation principles are truly intuitive to people and to which degree and under what circumstances they reflect human reasoning. By discovering the link between defined argumentation principles and human reasoning, we can bridge the gap between human reasoning and formal argumentation, allowing us to not only provide better explanations to end users but also increase their trust in AI applications.

In this project, we will first evaluate through user experiments and handcrafted examples whether people understand argumentation principles, if their reasoning follows these principles, and whether people’s satisfaction with specific principles affects their satisfaction with other principles.