Formal Encoding of Drama Ontology

Abstract

Abstract. The goal of this research is to lay the foundations for a formal theory of drama, that abstracts from the procedural and interactive aspects involved in the generation of dramatic content. Based on the structural accounts provided from traditional drama analysis, the theory proposed in this paper exploits an agent-based perspective on characters to provide a goal-based characterization of dramatic qualities. MotivationsSince the advent of digital media, character-based, narrative forms of communication have become commonplace in human-computer interaction, including user interfaces, entertainment, and education [1,2,3,4,5]. The need for autonomous behavior required by these applications has led scholars to adopt the agent techniques developed in artificial intelligence research to define and implement the virtual characters [6]. As applications have evolved into multi-character, interactive systems, the need for a centralized management of the plot execution has been realized by AI techniques for multi-agent coordination and cooperation [7,8].While the main effort of system developers has addressed the use of AI techniques in the production of interactive storytelling applications, designers have relied upon the widely acknowledged corpus of drama studies -from Aristotle's investigations to structuralist approaches -to characterize the dramatic qualities of virtual narratives. However, there is still lack for computational theory that exploits the conceptual tools of AI to characterize the principles of drama. The aim of this paper is to lay the foundations of a formal theory that systematizes the basic aspects of drama in a direct and explicit model, with an immediate integration with agent-based theories. The theory, called Drammar, abstracts from the procedural aspects of drama generation, and is intended as the starting point for specifying, implementing and evaluating practical storytelling systems in a principled way.Drammar is structured into two levels (see Figure 1). The actional level models the intentional behaviour of the characters in a plot by enforcing a BDI perspective on characters as intelligent, goal-directed agents: following Bratman's theory of practical reasoning [9], belief-desire-intention (BDI) agents form goals to pursue their desires, and, given their beliefs about the world, devise plans to achieve them. This level is augmented with a representation of emotions as provided by the OCC model [10]. The directional level accounts for the realization of a direction through the plot, by abstracting from the intentionality of the characters through the use of attributes that model the effect of plot incidents onto the characters' (i.e. agents') mental and emotional state.

Towards a semantics representation framework for narrative images

“…The studies on narrative models, which focus on describing the stories and show various relationships among the entities in the story, are helpful to the semantics representation of narrative images. The story contains the plots composed of the events and the entity behaviours; therefore, most narrative models, such as the drama ontology (Damiano et al , 2005) and the storytelling ontology (Nakasone and Ishizuka, 2006), establish the framework on the stories, events, agents and semantic roles. In recent years, many narrative models were proposed to describe the events in various fields, for example, the “simple event model” (SEM) proposed by the researchers of Vrije Universiteit Amsterdam (van Hage et al , 2011), the “Event Model F” proposed to represent the correlation between events in the real world (Scherp et al , 2009) and “Event Model E” created for multimedia content management (Westermann and Jain, 2007).…”

Mesotext : Digitised Emblems, Modelled Annotations and Humanities Scholarship

“…Bal (for the distinction between fabula, story and narrative text) and Campbell (for the analysis of the structure of heroic myths) are other theorists whose work is used in what is currently a very lively area of study (Gervás et al 2006;Swartjes & Theune 2006;Tuffield et al 2006). A formal model for drama is given by Damiano et al (2005) What these examples show is that modelling, in the sense of distinguishing and naming parts and describing their relations, is not something foreign to the study of literature or the wider humanities. For a model to be implemented on a computer, it may need a more explicit formulation, and of course the computer enforces consistency in its application (or at least exposes inconsistencies), but what the computer gives back is the possibility of manipulation.…”

A Fabula Model for Emergent Narrative

“…Much research has focused on understanding and formalizing event structures appearing in stories, from both the perspective of narratology (e.g., Propp's seminal work [8]), and story comprehension (e.g., [9]). Given the advantages of having an explicit narrative structure for story generation, it is surprising that -as also noted by [10] and [11] -story generation systems often lack a formalization of such structures. This was also the case for previous versions of the Virtual Storyteller, where the lack of a formal fabula representation resulted in a rather boring, literal account of the events occurring in our story world [1].…”

“…A recent attempt in formalizing drama structures is the Drammar ontology [10], a high-level ontology for the representation of drama on which our fabula structure can be said to elaborate. The ontology considers drama on two levels: the actional and the directional level, comparable to our idea of fabula and plot layers.…”