AI-RPG Toolkit: Towards A Deep Model Implementation for Improvisational Virtual Drama

The form of improvisational drama allows participants to have their own choices to influence the ongoing story, and each play results in a different ending. However, authoring such story contents requires ad hoc scripting, and static story structures lose ingenuity once users hacked through them. Our purpose is to develop a toolkit for: (1) fast authoring the story content, and (2) allow it for repeated plays yet retaining fresh interactive experience. While most similar applications have explicit, sophisticated story structures to ensure the number of possible interactions and endings in specific situations, we argue that characters should have enough background knowledge to make any improvisational choices. The more knowledge they have, the more sophisticated course of actions they may express. As a result, we take a deep-model approach to implement virtual agents, allowing them to deliberate and act with established knowledge in unexpected situations. Our methodology integrates the standardized web ontology language (OWL) from W3C into the game engine. OWL is utilized to define relations of concepts in descriptions of environments, e.g. the functionalities of objects, and personal traits in virtual dramas, e.g. motivation rules which map beliefs to desires in BDI agent structure. In a fashion of three-layered architecture which divides the system into agents, settings, and actual data, the character agents may infer their course of actions according to their strategies, sustained by settings defined in OWL to conceptualize the actual data. Defining background settings and possible storylines in OWL has many advantages, as OWL has many predefined classes of relations to build a well-structured knowledge base, and is portable to any other characters as long as they support this standard. As a result, the development of agent architecture and storylines can be separated as the situation when using scripts, but the knowledge created in OWL can be easily ported to any other applications effortlessly. The reuse of knowledge enables fast authoring of the story contents, and the accumulating background knowledge behind scenarios potentially allows characters to perform different behaviors. As this toolkit is still in an early stage of development, and knowledge base is far from exhaustive, it is sufficient for virtual agents to start an improvisational drama defined in OWL. Acknowledgments.