Social Practices for Social Driven Conversations in Serious Games
This work addresses a domain-specific problem for medical training, but it is incremental as it builds on an existing serious game and focuses on theoretical implementation.
The paper tackles the problem of training physicians in communicative skills by proposing a model of social practice as a theoretical framework to manage conversations in serious games, using a probabilistic model for selecting social practices to implement an agent architecture.
This paper describes the model of social practice as a theoretical framework to manage conversation with the specific goal of training physicians in communicative skills. To this aim, the domain reasoner that manages the conversation in the Communicate! \cite{jeuring} serious game is taken as a basis. Because the choice of a specific Social Practice to follow in a situation is non-trivial we use a probabilistic model for the selection of social practices as a step toward the implementation of an agent architecture compliant with the social practice model.