Information Technology Reference
In-Depth Information
Conceptual Model and System for Genre-Focused
Interactive Storytelling
Börje F. Karlsson
1
and Antonio L. Furtado
2
1
Microsoft Research, Beijing, China
2
Departamento de Informática, PUC-Rio, Rio de Janeiro, RJ, Brazil
borjekar@microsoft.com, furtado@inf.puc-rio.br
Abstract.
This paper describes a conceptual model for the definition of a genre
in the context of Interactive Storytelling and its implementation in LogTell-R, a
system for the interactive creation of stories. This work builds on a previous
system and experiments with plan recognition and discusses the foundations of
our model to allow the creation of varied and coherent stories within a genre.
Keywords:
storytelling, conceptual model, genre, plan generation / recognition.
1
Introduction
Schank [17] writes that humans understand the world in terms of stories. Arguably,
the human brain has a natural affinity not only for enjoying narratives, but also for
creating them [19]. Thus, the dynamic and interactive generation of stories is an
interesting problem. However, the generation of stories that are coherent and
interesting, while accounting for authors, and audience interactions remains an
unsettled issue. A wide range of approaches has tried to integrate storytelling and
interactive entertainment (narrative and gameplay), with limited success, mostly by
overly restricting user interaction options. While this may be a pragmatic attempt to
simplify the problem space, we feel that more generative approaches are required if
the field is to become truly successful. One of the main open problems in the
generative craft of stories resides in how stories may be generated and told. While it is
probably not possible to define an ideal general model for good stories, we claim that
sound methods to organize and combine events must be considered to confer enough
dramatic power to narratives. Some approaches do make use of special narrative
functions that enforce narrative principles in the context of the whole story (like
raising tension, introducing dilemmas, or some mechanism to “move the story
forward”) [10], but while these efforts improve user engagement they still do not
guarantee diversity of experience. A possible avenue for the development of more
generative systems that can create interesting stories is related to the
definition of a
genre
. Once a genre is specified with some rigour in a constructive way, it becomes
possible not only to determine whether a given plot is a legitimate representative of
the genre, but also to generate such plots. To tackle the problem of creating a
more generative model to support the production of stories, we draw on previous
