IDTension
Template:Under construction Main page - WP1: Narrative Formalisms - IS Systems - IDTension
Contents
Availability
Successive versions of IDtension have been demoed in academic conferences since 2001.
A beta version is accessible here (link not yet distributed)
A "released" version should arrive during 2009.
Technical Description
IDtension is an Interactive Drama engine where the narrative unfolds as the user decides what actions the main character will perform in relation to the other characters in the story.
Fully developed in Java, IDtension's narrative management is characterized by the following distinctive features:
- Atemporal structure: A story is not described as more or less pre-authored “chunks” of stories such as scenes but as an atemporal structure of goals, tasks, obstacles and values. This structure resembles a classical goal driven architecture for agents, but
- It is described in narrative terms (obstacles, ethical values).
- It is not attached to a character but centralized in a "state of the world"
- Second order actions: In order to obtain a large number of possible actions without demultiplying the authoring effort, actions are described according to a second order formalism. For example: inform(x,y,goal(z,g,u)) means that a character x inform another character y that a third character z has a given goal g, with a fourth character u as a parameter. It could produce the following dialog line: Mary to John: Did you know that Bill wished to be loved by Rachel? In the formula above, “inform”, and “goal” are hardcoded, while x, y, z, u and g are variables that can take any authored-defined values. Hardcoded elements correspond to narratology-inspired fundamental narrative actions and states. This mechanism ensures a large number of produced actions with a limited authored material.
- Model of the player: The action selection mechanism for Non Playing Characters (NPC) is based on a general model of the player’s perception. Indeed, each action is evaluated according to a set of narrative criteria. While some roughly correspond to agent’s rationality (ethical consistence, motivational consistence), others are narratively motivated, such as “conflict”.
Result Description (end user perspective)
The latest demo consists of a playable interactive drama called « The Mutiny », written in collaboration with Olivier Marty. The player is one of the characters in the story, a sailor jailed in a 17th century galleon with three other prisoners, after a failed plunder against the ship. His goal is now to take the leadership of the galleon by preparing a small riot (only four prisoners...). This story enables a high degree of interactivity. The user is given many possibilities of actions such as ask/trade/rob/rob armed with a knife an object, ask another character about his/her taste (in order to offer him/her the object he/she likes), try to flatter other characters to have allies, etc. Furthermore, the user is involved in dialogs in which he can refer to the above actions, for example, inform that he can/want to perform these actions or that he/she has failed/succeeded to perform them.
This interactive drama is text-based, although a 3D version has also been implemented.
Strong Points
The range of actions is higher than in other Interactive Drama systems. In the middle of The Mutiny (see above), more than 100 choices can be found, which are interpreted by the system.
Limitations
The cohesion of the story is not always maintained. For example, some user's dialog actions do not get any answers from non player characters.
The authoring of stories is difficult.
Main Publications
- Szilas, N., 1999. Interactive Drama on Computer: Beyond Linear Narrative. In Proc. AAAI Fall Symposium on Narrative Intelligence (North Falmouth MA). Menlo Park: AAAI Press. [1]
The first paper on IDtension. - Szilas, N., 2003. IDtension: a narrative engine for Interactive Drama. In Göbel et al. (eds) Proc. TIDSE’03. Frauenhofer IRB Verlag. [2]
General presentation of IDtension, with two experimental simulations on a simple story: one automatic generation of stories, one interactive simulation. - Szilas, N.， 2007. A Computational Model of an Intelligent Narrator for Interactive Narratives. Applied Artificial Intelligence, 21(8), pp. 753-801, Sept. 2007. [3]
The most comprehensive presentation of the computational model.
Supporting Narrative Theories
link to the narrative theory underlying the story created by the tool (WP1)
Computational Model
link to the description of the computational model used within the tool (WP1)
