핵심 개념
This work introduces a semantic game and a provability game for reasoning about group polarization using the modal logic PNL. The games lead to the development of cut-free sequent calculi for PNL and its extensions, which can be used for automated reasoning about social network dynamics and polarization.
초록
The paper focuses on formal reasoning about group polarization, where individuals become more extreme in their opinions after interacting within a group. The authors use the modal logic PNL, which captures the notion of agents agreeing or disagreeing on a given topic, as the basis for their work.
The key contributions are:
Introducing a semantic game that characterizes truth in a given PNL model. This game provides an alternative to the standard Kripke semantics and supports dynamic reasoning about concrete network models.
Defining a disjunctive provability game that lifts the semantic game to logical validity. This game is shown to be adequate with respect to the logic, in the sense that the existence of winning strategies for the proponent corresponds to logical validity.
Developing cut-free sequent calculi DS and DScc for PNL and its extension to collectively connected models, respectively. These proof systems are derived from the structure of the provability game and provide a foundation for automated reasoning about group polarization.
Showing how the global and local link change modalities from previous work can be incorporated into the proposed framework, enabling the analysis of dynamic social network models.
The authors demonstrate the expressiveness of their approach through examples, and mention a prototypical implementation using rewriting logic and Maude.