A Computational Approach to Etiquette and Politeness
Keywords: social interaction behavior, etiquette, politeness, avatar, NPCs, cross cultural training, training games
Abstract: Characters or agents which react appropriately—-taking offense when reasonable, giving deference where appropriate, etc.-- are a fundamental need for believability and accuracy in simulations of social interactions (including culture-specific and multi-cultural interactions). This is especially true for applications where complex and realistic interactions with intelligent agents are important-- such as cross-cultural training for military personnel. We have developed a quantitative, computational implementation of a rich, universal theory of human-human “politeness” behaviors and the culture-specific interpretive frameworks for them (labeled “etiquette”) from sociology, linguistics and anthropology. This model links observable and inferred aspects of power and familiarity relationships, the degree of imposition of an act (each of which have implications for roles and intents) and the actor’s character to produce expectations about politeness behaviors. By using observations of politeness behaviors (or their lack), the same model permits inferences and updates about those attributes. We present the algorithm we have developed and describe its results in scoring the degree of politeness or rudeness across 8 test cases. We see applicability of this model to interactive agent behavior generation and adaptation through the creation of modular, cross-cultural etiquette libraries. While other methods of interactive behavior generation are available (e.g., behavior scripting) our modular, computational approach should provide substantial payoffs in terms of reducing software development costs and/orincreasing the breadth of an agent’s social interaction behaviors.