Academia.eduAcademia.edu

Outline

Title
Abstract
Figures
Introduction
Conclusion
References

User intervention in virtual interactive storytelling

Profile image of Marc CavazzaMarc Cavazza

2001

Abstract

Abstract: In this paper, we describe how users of virtual storytelling environments can interact with the artificial actors by interfering with their plans, which determine plot orientation. We have developed a fully implemented prototype based on the UnrealTM game engine and carried experiments with a simple sitcom-like scenario. We first present our storyline, and follow previous work on AI planning formalisms to describe the techniques used for the implementation of the dynamic narrative.

Figures (10)
Figure 1: Unsolved “Take her out” AND/OR graph (showing user and NLP intervention levels) fee", SEER: Wenn IN A ean Re eer eTT AeA res “UNM Ea Nee Oe em eM ts CeO The scenario presented in this paper is based on a sitcom-like episode. The main focus of the scenario is for the principal character to invite the main female character on a date. The male character, “Ross”, must find ways of succeeding by acquiring personal information about the female character, “Rachel”, and then will need to isolate Rachel in order to talk to her privately.
Figure 1: Unsolved “Take her out” AND/OR graph (showing user and NLP intervention levels) fee", SEER: Wenn IN A ean Re eer eTT AeA res “UNM Ea Nee Oe em eM ts CeO The scenario presented in this paper is based on a sitcom-like episode. The main focus of the scenario is for the principal character to invite the main female character on a date. The male character, “Ross”, must find ways of succeeding by acquiring personal information about the female character, “Rachel”, and then will need to isolate Rachel in order to talk to her privately.
Figure 2: The virtual set Sub-goals describe the partial plans that need to be successfully completed to progress to a full solution’. The main character needs to first gather details regarding the female character’s availability, her tastes in restaurant, movies, theatre or other personal information. To acquire information about Rachel, three possible routes are provided that can be represented as sub-goals of the top-level goal. Ross may go and borrow her diary without her knowing, or borrow her personal digital assistant (PDA), or go and talk to her best friend. The background knowledge acquired will provide the main character with further details on the current situation.
Figure 2: The virtual set Sub-goals describe the partial plans that need to be successfully completed to progress to a full solution’. The main character needs to first gather details regarding the female character’s availability, her tastes in restaurant, movies, theatre or other personal information. To acquire information about Rachel, three possible routes are provided that can be represented as sub-goals of the top-level goal. Ross may go and borrow her diary without her knowing, or borrow her personal digital assistant (PDA), or go and talk to her best friend. The background knowledge acquired will provide the main character with further details on the current situation.
For instance, according to the static heuristic values that were attributed to the sub-goals, the best partial solution plan, at the time of the first computation, would be for Ross to borrow Rachel’s diary (“Borrow her diary” sub-goal). The termina actions corresponding to this sub-goal are to reach the diary, pick it up and retrieve the relevant information. The artificial actor manages to reach the location of the object, but the object is not available anymore because the user intervened with the virtua environment resources, by stealing the diary (figure 3). As the terminal actions are processed sequentially their outcome is important to the dynamic search. Al the terminal actions for the same sub-goal must be validated for the partial sub-plan to be solved. As shown in figure 5, the first partial solution fails when the artificial actor is unable to collect the diary. Figure 3: User interference by “stealing the diary”
For instance, according to the static heuristic values that were attributed to the sub-goals, the best partial solution plan, at the time of the first computation, would be for Ross to borrow Rachel’s diary (“Borrow her diary” sub-goal). The termina actions corresponding to this sub-goal are to reach the diary, pick it up and retrieve the relevant information. The artificial actor manages to reach the location of the object, but the object is not available anymore because the user intervened with the virtua environment resources, by stealing the diary (figure 3). As the terminal actions are processed sequentially their outcome is important to the dynamic search. Al the terminal actions for the same sub-goal must be validated for the partial sub-plan to be solved. As shown in figure 5, the first partial solution fails when the artificial actor is unable to collect the diary. Figure 3: User interference by “stealing the diary”
Figure 4: Re-planning after action failure
Figure 4: Re-planning after action failure
Figure 6: Terminal action performed to validate a sub-plan. At this stage, the search process is reactivated from the upper level sub-goal (“Acquire information’) in order to find a new best partial solution plan. The AO*-based search results in the choice of the sub-goal “Borrow her PDA”. This time, all the terminal actions are successful (figure 5, 6), hence validating the current sub-plan as the best partial solution.
Figure 6: Terminal action performed to validate a sub-plan. At this stage, the search process is reactivated from the upper level sub-goal (“Acquire information’) in order to find a new best partial solution plan. The AO*-based search results in the choice of the sub-goal “Borrow her PDA”. This time, all the terminal actions are successful (figure 5, 6), hence validating the current sub-plan as the best partial solution.
Figure 5: Interleaving planning and action (sub-graph)
Figure 5: Interleaving planning and action (sub-graph)
We have developed a Windows application (figure 7) in C++, based on the Ear™ SDK*. Ear™ recognises language input defined by a Context-Free Grammar (CFG) and allows several forms of optional entries and “garbage models” that make it possible to implement multi-keyword spotting. The grammar for the recognised sub-language is actually entered in a specific file as a transition network. These are constructed using a markup-like language tags in a lexicon editor (figure 8).
We have developed a Windows application (figure 7) in C++, based on the Ear™ SDK*. Ear™ recognises language input defined by a Context-Free Grammar (CFG) and allows several forms of optional entries and “garbage models” that make it possible to implement multi-keyword spotting. The grammar for the recognised sub-language is actually entered in a specific file as a transition network. These are constructed using a markup-like language tags in a lexicon editor (figure 8).
Figure 8: Context-Free Grammar NII INP II IOI IIE IAI ILA YIN NI NINA III EON NE EE We have __s previously implementations of natura virtual actors [Cavazza and integrated syntactic and of using multiple speech recognition paradigms, the Considering the possibility accommodate various ungrammatical ones prod is still necessary to recover Pa described several language interfaces for mer, 1999], based on semantic _ parsing. parsing step should entries including uced by multi-keyword spotting (i.e., plain keyword seq uences, from which it per semantic relations pro between keywords). We have hence adapted our parser to be able to work with a wider range of input: its current version produces partial parses and merges the resulting semantic structures on a content basis.
Figure 8: Context-Free Grammar NII INP II IOI IIE IAI ILA YIN NI NINA III EON NE EE We have __s previously implementations of natura virtual actors [Cavazza and integrated syntactic and of using multiple speech recognition paradigms, the Considering the possibility accommodate various ungrammatical ones prod is still necessary to recover Pa described several language interfaces for mer, 1999], based on semantic _ parsing. parsing step should entries including uced by multi-keyword spotting (i.e., plain keyword seq uences, from which it per semantic relations pro between keywords). We have hence adapted our parser to be able to work with a wider range of input: its current version produces partial parses and merges the resulting semantic structures on a content basis.
Figure 9: Targets for user intervention
Figure 9: Targets for user intervention

Related papers

Interactive Digital Storytelling: Automatic Direction of Virtual Environments
Federico Peinado

Retrieved October, 2006

Virtual environments are a digital media in which each time more stories are told with the purpose of teaching, training, entertaining or communicating something to the users. Predefined scripts, plenty of interaction freedom and even the simulation of a fictional world and its inhabitants are not enough to create quality stories. Since late eighties, a growing scientific community has being developing intelligent systems able to direct automatically everything that happens in a virtual environment, in order to adapt the story to the application goals and the particular characteristics of each interactor. This paper presents an updated review of the steps forward, results in the form of implemented systems and pending tasks of this research line.

View PDFchevron_right
DIRACT: Agent-based Interactive Storytelling
Yundong Cai

2010

A lot of researches haven been done on the interactive storytelling authoring , e.g. by a director agent, or by interactions among a number of character agents. However, it is still difficult to construct the interactive storytelling for novice users, due to a need of various agents development and complex communication among the agents. We propose an agent-based interactive storytelling architecture, namely DIRACT (short of "Direct and Act"). It is composed of numerous atomic DIRACT agents, which are goal oriented and use an unified communication protocol. By removing the difference between the director and character, each DIRACT agent can either direct or act the story in real-time. As a result, it is scalable to story complexity, reusable for inheritance and makes the interactive storytelling authoring much easier.

View PDFchevron_right
Agent-Oriented methodology for interactive storytelling (AOMIS)
Yundong Cai

2011

Interactive virtual storytelling attracts a lot of research interests to create engaging, interactive and dynamic storytelling. However, currently there is a big gap between story plot modeling to believable character modeling and it is still a great challenge to create an interactive storytelling system from text-based stories for novice. In this paper, we propose an Agent-Oriented Methodology for Interactive Storytelling (AOMIS) to guide novice to generate interactive storytelling system. In this methodology, interactive storytelling system is constructed as a multi-agent system of director agent and character agents. The interactive storytelling is simulated as the goal pursuing of the agents affecting by user interactions and context changes. By viewing the interactive storytelling in the perspective of goal-oriented agents, our methodology is a hybrid of plot-oriented interactive storytelling and character-based interactive storytelling.

View PDFchevron_right
Managing interaction between users and agents in a multi-agent storytelling environment
R. Michael Young

2003

Abstract This paper describes an approach for managing the interaction of human users with computer-controlled agents in an interactive narrative-oriented virtual environment. In these kinds of systems, the freedom of the user to perform whatever action she desires must be balanced with the preservation of the storyline used to control the system's characters. We describe a technique, narrative mediation, that exploits a plan-based model of narrative structure to manage and respond to users' actions inside a virtual world.

View PDFchevron_right
Character-driven story generation in interactive storytelling
Marc Cavazza

2001

Abstract We describe a fully implemented prototype for interactive storytelling using the Unreal TM engine. We describe the important mechanisms involved in the variability of plot instantiations, within a scenario of sitcom genre. We also provide an evaluation of the concepts of how the dynamic interactions between agents and/or the user influence the generation of story, with first results of examples

View PDFchevron_right
Distributed interactive narrative planning system
R. Michael Young

2006

Abstract (Riedl, Saretto, & Young 2003) describes narrative mediation, a plan-based technique for addressing human user actions that conflict with the story structure maintained by computer-controlled agents within interactive narrative environments. This technique is used to generate real-time system responses to unanticipated user actions while balancing user freedom with story coherence. Narrative mediation has previously only been used for a single plan representing an entire story.

View PDFchevron_right
A new approach to interactive drama: From intelligent characters to an intelligent virtual narrator
Nicolas Szilas

Proceedings of the AAAI Spring Symposium on AI and …, 2001

Goal of this research We aims at building a new genre of artistic expression, called interactive drama. An interactive drama is a drama (a narrative where the story is told by directly showing actions to the audience) where audience can act upon some events in the story. In this ...

View PDFchevron_right
Character-based interactive storytelling
Vinit Agarwal

Intelligent Systems, IEEE, 2002

View PDFchevron_right
Virtual Actors’ Behaviour for 3D Interactive Storytelling
Marc Cavazza

2001

Abstract In this paper, we describe a method for implementing intelligent behaviour for artificial actors in the context of interactive storytelling. We have developed a fully implemented prototype based on the Unreal™ game engine and carried experiments with a simple sitcom-like scenario. We discuss the central role of artificial actors in interactive storytelling and how real-time generation of their behaviour participates to the creation of a dynamic storyline.

View PDFchevron_right
A Study of Interactive Narrative from User's Perspective
David Milam

Handbook of Multimedia for Digital …, 2009

The topic of interactive narrative has been under research for many years. Many regard the term itself as an oxymoron [1] [2], while others see narrative as an integral part of every interactive production [3] . While there has been much research exploring the development of new algorithms that enable and enhance interactive narratives, there has been little research focusing on the question of how players understand and internalize their interactive narrative experiences. In this chapter, we discuss the interactive narrative experience as seen through the users" eyes. Specifically, we report on a study we conducted using a phenomenological approach to explore the participants" lived experience of playing an interactive narrative. We chose to use Façade, as an example interactive narrative for our study due to its accessibility and its focus on social relationships, conflict, and drama as its core mechanics. As we look into the experience of interactive narrative, we reflect on enhancements to the design of future interactive narratives and discuss open problems in the field of interactive narrative.

View PDFchevron_right

References (10)

  1. J. Altman, and R. Nakatsu. Interactive Movies: Techniques, Technology and Content. Course Notes n. 16, ACM SIGGRAPH'97, 1997.
  2. M. Cavazza and I. Palmer. Natural Language Control of Interactive 3D Animation and Computer Games. Virtual Reality, 4:85-102; 1999.
  3. K Dautenhahn : Story-Telling in Virtual Environments, Working Notes: Intelligent Virtual Environments, Workshop at the 13 th Biennial European Conference on Artificial Intelligence (ECAI-98), Brighton, UK, 1998.
  4. E. A. Hansen, S. Zilberstein, V. A. Danilchenko, Anytime Heuristic Search: First Results, CMPSCI Technical Report 97-50, 1997.
  5. K. Knight and E. Rich. Artificial Intelligence 2 nd Edition. McGraw Hill, 1991.
  6. N.J. Nilsson. Principles of Artificial Intelligence. Palo Alto, CA. Tioga Publishing Company, 1980.
  7. J. Pearl. Heuristics: Intelligent Search Strategies for Computer Problem Solving. Reading (Massachusetts), Addison-Wesley, 1984.
  8. N. M Sgouros, G. Papakonstantinou, P. Tsanakas, A Framework for Plot Control in Interactive Story Systems, in Proc. AAAI-96, Portland OR, AAAI Press, 1996.
  9. B. L. Webber, N. I. Badler, B. D. Eugenio, C. Geib, L. Levison, M. Moore, Instructions, Intentions and Expectations, IRCS Technical Report 94-01, 1994.
  10. R.M. Young. Creating Interactive Narrative Structures: The Potential for AI Approaches. AAAI Spring Symposium in Artificial Intelligence and Computer Games, AAAI Press, 2000.

Related papers

Planning characters' behaviour in interactive storytelling
Marc Cavazza

2002

Abstract In this paper, we describe a method for implementing the behaviour of artificial actors in the context of interactive storytelling. We have developed a fully implemented prototype based on the Unreal Tournament™ game engine, and carried experiments with a simple sitcom-like scenario. We discuss the central role of artificial actors in interactive storytelling and how real-time generation of their behaviour participates in the creation of a dynamic storyline.

View PDFchevron_right
Characters in search of an author: AI-based virtual storytelling
Marc Cavazza

2001

In this paper, we present the first results obtained with an interactive storytelling prototype. Our main objective is to develop flexible character-based systems, which nevertheless rely on narrative formalisms and representations. Characters' behaviours are generated from plan-based representations, whose content is derived from narrative formalisms.

View PDFchevron_right
Agents’ interaction in virtual storytelling
Marc Cavazza

2001

In this paper we describe a fully implemented prototype for interactive storytelling using the Unreal™ engine. Using a sit-com like scenario as an example of how the dynamic interactions between agents and/or the user dramatise the emerging story. Hierarchical Task Networks (HTNs) are formalised using AND/OR graphs, which are used to describe the many possible variations of the story at a sub-goal level, and the set of all behaviours (from a narrative perspective) of the primary actors at a terminal action level.

View PDFchevron_right
Behavioural Interaction of Characters for Virtual Storytelling
Marc Cavazza

2001

ABSTRACT In this paper we describe a fully implemented prototype for interactive storytelling using the Unreal™ engine. We describe the important mechanisms involved in the variability of plot instantiations, within a scenario of sitcom genre. We also provide an evaluation of the concepts of how the dynamic interactions between agents and/or the user influence the generation of story, with first results of examples

View PDFchevron_right
Towards an architecture for intelligent control of narrative in interactive virtual worlds
R. Michael Young

2003

Abstract The creation of novel, engaging and dynamic interactive stories presents a unique challenge to the designers of systems for interactive entertainment, education and training. Unlike conventional narrative media, an interactive narrative-based system may be required to generate its own story structure, determine the appropriate interface elements to use to convey the storys action and manage the effective interaction of a user within the story as it plays out.

View PDFchevron_right
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved