scispace - formally typeset
Open AccessJournal ArticleDOI

A Computational Model of Narrative Generation for Surprise Arousal

Reads0
Chats0
TLDR
The results of the experiments show strong support that Prevoyant effectively generates a discourse structure for surprise arousal in narrative, and a methodology to evaluate surprise in narrative generation using a planning-based approach based on the cognitive model of surprise causes.
Abstract
This paper describes our effort for a planning-based computational model of narrative generation that is designed to elicit surprise in the reader's mind, making use of two temporal narrative devices: flashback and foreshadowing. In our computational model, flashback provides a backstory to explain what causes a surprising outcome, while foreshadowing gives hints about the surprise before it occurs. Here, we present Prevoyant, a planning-based computational model of surprise arousal in narrative generation, and analyze the effectiveness of Prevoyant. The work here also presents a methodology to evaluate surprise in narrative generation using a planning-based approach based on the cognitive model of surprise causes. The results of the experiments that we conducted show strong support that Prevoyant effectively generates a discourse structure for surprise arousal in narrative.

read more

Content maybe subject to copyright    Report

ABSTRACT
BAE, BYUNG CHULL. A Computational Model of Narrative Generation for Surprise
Arousal. (Under the direction of Professor R. Michael Young).
This dissertation describes work to develop a planning-based computational model of
narrative generation designed to elicit surprise in the mind of a reader. To this end, my
approach makes use of two narrative devices flashback and foreshadowing. While surprise
plays an important role for attention focusing, learning, and creativity, little effort has been
made to build a computational framework for surprise arousal in narrative. In my
computational model, flashback provides a backstory to explain what causes a surprising
outcome, while foreshadowing gives hints about the surprise before it occurs. In this work I
focus on the arousal of surprise emotion as a cognitive response which is based on a reader's
cognitive appraisal of a given situation. In this dissertation I present Prevoyant, a planning-
based computational model of surprise arousal in narrative generation, and analyze the
effectiveness of Prevoyant. To build a computational model of the unexpectedness in surprise,
I adopt a cognitive model of surprise based on expectation failure.
There are two contributions made by this dissertation. First, I present a computational
framework for narrative generation designed to elicit surprise. The approach makes use of a
two-tier model of narrative and draws on Structural Affect Theory, which claims that a
reader‟s emotions such as surprise or suspense are closely related to narrative structure.
Second, I present a methodology to evaluate surprise in narrative generation using a
planning-based approach based on the cognitive model of surprise causes. The results of the
experiments that I conducted show strong support that my system effectively generates a
discourse structure for surprise arousal in narrative.

A Computational Model of Narrative Generation for Surprise Arousal
by
Byung Chull Bae
A dissertation submitted to the Graduate Faculty of
North Carolina State University
In partial fulfillment of the
Requirements for the degree of
Doctor of Philosophy
Computer Science
Raleigh, North Carolina
June 26, 2009
APPROVED BY:
_______________________________ ______________________________
Dr. James C. Lester Dr. Brad Mehlenbacher
_______________________________ ______________________________
Dr. Robert Rodman Dr. R. Michael Young
Chair of Advisory Committee

ii
DEDICATION
To
my parents,
my wife Yuna,
and
my precious daughter Hyunji (Iris)

iii
BIOGRAPHY
Byung Chull Bae grew up in Seoul, South Korea, where he developed his love of music,
stories, and movies. He attended Korea University majoring in Electronics Engineering,
which provided the curriculums of introductory Computer Science and advanced Computer
Engineering. He earned his Bachelor's Degree in Electronics Engineering in 1993 and then
continued his graduate study under the advisement of Dr. Duk-Jin Kim at Korea University.
He earned his Master's Degree in Electronics Engineering in 1998 with his thesis, A Study on
Internetworking between PSTN and B-ISDN, which presented a new way of internetworking
technique for the reliable data transition between different networks.
From 1998 to 2002, he worked at LG Electronics R&D Center as an assistant researcher,
building various hardware and software programs for mobile networking environments. In
December 2002, he got married to Yun Gyung Cheong, who was a Ph.D. student in
Computer Science at North Carolina State University at that time. In August 2003, he moved
to the United States and entered the Master/Ph.D. program in Computer Science at North
Carolina State University. Soon he joined the Liquid Narrative group led by Dr. R. Michael
Young and performed research in the area of automated background music generation for
narrative. In 2005, he earned his Master‟s degree in Computer Science and continued his
study on emotion, cognitive science, and narrative. In March 2007, he became a father of a
beautiful baby girl, Hyunji Iris Bae. In November 2008, he won the Best Paper Award at the
first International Conference on Interactive Digital Storytelling. After graduation, he plans
to pursue a career in research on interactive storytelling and emotions of intelligent agents.

iv
ACKNOWLEDGEMENTS
Foremost, I thank my committee, Dr. James Lester, Dr. Brad Mehlenbacher, and Dr. Robert
Rodman for their time and comments. I thank R. Michael Young, my thesis advisor for his
warm but strict guidance through the process of finishing this dissertation. I wouldn't be here
where I am without him. Thanks for spending his time, effort, and patience to keep me on the
right track.
I thank Dr. Marie-Laure Ryan for her kind and encouraging words about my work. I
thank Dr. Jon Doyle, Dr. Carla Savage, Dr. Robert St. Amant, and other Computer Science
professors at NC State for their enthusiastic class. I learned invaluable lessons from them. I
thank Dr. David Thuente, Ms. Margery Page, and other Computer Science staffs at NC State
University. I thank students in Computer Science and at NC State who voluntarily
participated in my experimental studies.
I thank my fellow students at NC State University, Pat Cash, David Christian, Mike
Dickheiser, Michael Dominguez, Julius Goth, Joseph Grafsgaard, Oliver Gray, Eun-Young
Ha, Justin Harris, Dr. Arnav Jhala, Seung Yong Lee, Dr. Sunyoung Lee, Sam Munilla, Dr.
Brad Mott, Dr. James Niehaus, Dr. Mark Riedl, Stephen Roller, Jonathan Rowe, Brian Shiver,
Jim Thomas, Tommy Vernieri, Kevin Vaughan, and Joe Winegarden for enjoyable
discussion on many topics. I enjoyed the six years at NC State because of you guys.
I thank Korea University alumni attending (or attended) at NC State who gave me good
support and strength for six years. Special thanks for Seung-Jun Shin for his statistical
counseling regarding my study.
I thank Chang-Yeol Lee for his good friendship over 25 years. I thank Jeong-Im Yom for
her timeless friendship over 20 years and her family for their hospitality. I thank Pat and Ron
Leith for their warmness and inviting my family to their Thanksgiving Day party every year.
I also thank Andy and Cheryl White for their kindness. I thank Mike, the owner of Global
Village coffee house on Hillsborough Street, for his gentle manner and good coffee. With
Yuna, my wife, I spent a lot of good time there reading numerous papers. I also thank the

Citations
More filters
Journal ArticleDOI

A Survey on Story Generation Techniques for Authoring Computational Narratives

TL;DR: By examining the successes of existing research, this survey identifies potential new research directions in the field of computational narrative and identifies the advantages of developing a standard model of narrative to allow for collaboration between plot and space automation techniques.
Journal ArticleDOI

Exploring the Development of ‘Interest’ in Learning English as a Foreign/Second Language:

TL;DR: In this paper, a new conceptual perspective for understanding students' engagement in L2 learning, in particular the concept of interest development in learning English, was proposed, and the authors investigated situational features that trigger interest in learning.
Proceedings Article

Evaluating the Pairwise Event Salience Hypothesis in Indexter.

TL;DR: A crowd-sourced study of 24 short text stories that control for content, text, and length supports the Pairwise Event Salience Hypothesis and motivates the development of a richer model that accounts for intervening events, narrative complexity, and episodic memory decay.
Journal ArticleDOI

The campfire storytelling system – automatic creation and modification of a narrative

TL;DR: The results of these evaluations suggest that Campfire STS is able to create interesting and novel stories, as well as change them seamlessly in mid-stream while still maintaining interesting and entertaining stories.
References
More filters
Book

The Cognitive Structure of Emotions

TL;DR: In this paper, a cognitive theory of emotion is proposed, which describes the organization of emotion types and the implications of the emotions-as-valenced-reactions claim, and the boundaries of the theory Emotion words and cross-cultural issues.
Journal ArticleDOI

Narrative discourse : an essay in method

TL;DR: Cutler as mentioned in this paper presents a Translator's Preface Preface and Preface for English-to-Arabic Translating Translators (TSPT) with a preface by Jonathan Cutler.
Journal ArticleDOI

Towards a Cognitive Theory of Emotions

TL;DR: In this paper, a theory is proposed that emotions are cognitively based states which co-ordinate quasi-autonomous processes in the nervous system, and that complex emotions are derived from a small number of basic emotions and arise at junctures of social plans.
Related Papers (5)
Frequently Asked Questions (12)
Q1. What are the contributions in this paper?

This dissertation describes work to develop a planning-based computational model of narrative generation designed to elicit surprise in the mind of a reader. In this work I focus on the arousal of surprise emotion as a cognitive response which is based on a reader 's cognitive appraisal of a given situation. The approach makes use of a two-tier model of narrative and draws on Structural Affect Theory, which claims that a reader‟s emotions such as surprise or suspense are closely related to narrative structure. 

As future work, I consider three potential extensions of Prevoyant: a story plan‟s cinematic realization ; combination of surprise and suspense ; and interactive storytelling. The following sections explain these three areas for future work. 

Universe has a nonlinear narrative space structure because of its boundless extensibility using the concept of plot fragment and character development, which provides central ideas with later nonlinear interactive systems. 

In terms of entertainment, surprise has a functional role of maintaining and focusing attention, which helps to keep the audience from distraction. 

Three temporal narrative devices particularly associated with cinematic narratives (Chatman, 1978; Prince, 2003; Bordwell, 1986) are often used by storytellers to manipulate3the presentation order of story events: flashforward, foreshadowing, and flashback. 

In narratives, one of the main functional roles of surprise is to stimulate a reader‟s cognitive interest, which can be drawn out from the narrative structure rather than the emotional impact of the story (Kintsch, 1980). 

The selection of the flashback with high causal importance contributes to the story interestingness, increasing the postdictability in retrospect. 

In order to simulate a user‟s response to an interactive drama in the virtual environments, they conducted “first live interactive improvisation” experiments with human actors and a drama director. 

Both Erica‟s rather ambiguous identity and the lack of Jack‟s activity as the antagonist in the story were pointed out as reasons for excluding these characters from subjects‟ favorites. 

Unanticipated incongruities are passive cognitive activity in that the agent does not actively expect any propositions in conflict with an input proposition. 

75A one-way ANOVA confirmed that at least two discourse types had significant effects on the ratings of surprise, F(2, 15 df) = 9.61, p = .002, and on the ratings of interestingness, F(2, 15 df) = 5.24, p = .019. 

An excellent example of incongruity resolution using flashback would be the flashback scene at the ending in The Usual Suspects (1995), where the detective interrogating “Verbal” Kint realizes the truth – who Keyser Söze actually is – right after releasing Kint.