Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks

Brent Harrison; Christopher Purdy; Mark O. Riedl

Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks

Brent Harrison, Christopher Purdy, Mark O. Riedl

Research output: Contribution to conference › Paper › peer-review

14 Scopus citations

Abstract

In this paper, we introduce an approach to automated story generation using Markov Chain Monte Carlo (MCMC) sampling. This approach uses a sampling algorithm based on Metropolis-Hastings to generate a probability distribution which can be used to generate stories via random sampling that adhere to criteria learned by recurrent neural networks. We show the applicability of our technique through a case study where we generate novel stories using an acceptance criteria learned from a set of movie plots taken from Wikipedia. This study shows that stories generated using this approach adhere to this criteria 85%-86% of the time.

Original language	English
Pages	191-197
Number of pages	7
State	Published - 2017
Event	13th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, AIIDE 2017 - Snowbird, Little Cottonwood Canyon, United States Duration: Oct 5 2017 → Oct 9 2017

Conference

Conference	13th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, AIIDE 2017
Country/Territory	United States
City	Snowbird, Little Cottonwood Canyon
Period	10/5/17 → 10/9/17

Bibliographical note

Publisher Copyright:
Copyright © 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.

Funding

This work was supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No. W911NF-15-C-0246 and by the National Science Foundation under Grant No. IIS-1350339.

Funders	Funder number
National Science Foundation Arctic Social Science Program	IIS-1350339
National Science Foundation Arctic Social Science Program
Defense Advanced Research Projects Agency	W911NF-15-C-0246
Defense Advanced Research Projects Agency

ASJC Scopus subject areas

General Engineering
Software
Artificial Intelligence
Human-Computer Interaction
Computer Science Applications
Computer Graphics and Computer-Aided Design

Cite this

@conference{f344fec8b1a340e5bcaf7b2a7158d845,

title = "Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks",

abstract = "In this paper, we introduce an approach to automated story generation using Markov Chain Monte Carlo (MCMC) sampling. This approach uses a sampling algorithm based on Metropolis-Hastings to generate a probability distribution which can be used to generate stories via random sampling that adhere to criteria learned by recurrent neural networks. We show the applicability of our technique through a case study where we generate novel stories using an acceptance criteria learned from a set of movie plots taken from Wikipedia. This study shows that stories generated using this approach adhere to this criteria 85%-86% of the time.",

author = "Brent Harrison and Christopher Purdy and Riedl, {Mark O.}",

note = "Publisher Copyright: Copyright {\textcopyright} 2017, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.; 13th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, AIIDE 2017 ; Conference date: 05-10-2017 Through 09-10-2017",

year = "2017",

language = "English",

pages = "191--197",

}

Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks. / Harrison, Brent; Purdy, Christopher; Riedl, Mark O.
2017. 191-197 Paper presented at 13th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, AIIDE 2017, Snowbird, Little Cottonwood Canyon, United States.

Research output: Contribution to conference › Paper › peer-review

TY - CONF

T1 - Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks

AU - Harrison, Brent

AU - Purdy, Christopher

AU - Riedl, Mark O.

PY - 2017

Y1 - 2017

N2 - In this paper, we introduce an approach to automated story generation using Markov Chain Monte Carlo (MCMC) sampling. This approach uses a sampling algorithm based on Metropolis-Hastings to generate a probability distribution which can be used to generate stories via random sampling that adhere to criteria learned by recurrent neural networks. We show the applicability of our technique through a case study where we generate novel stories using an acceptance criteria learned from a set of movie plots taken from Wikipedia. This study shows that stories generated using this approach adhere to this criteria 85%-86% of the time.

AB - In this paper, we introduce an approach to automated story generation using Markov Chain Monte Carlo (MCMC) sampling. This approach uses a sampling algorithm based on Metropolis-Hastings to generate a probability distribution which can be used to generate stories via random sampling that adhere to criteria learned by recurrent neural networks. We show the applicability of our technique through a case study where we generate novel stories using an acceptance criteria learned from a set of movie plots taken from Wikipedia. This study shows that stories generated using this approach adhere to this criteria 85%-86% of the time.

UR - http://www.scopus.com/inward/record.url?scp=85165725341&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85165725341&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85165725341

SP - 191

EP - 197

T2 - 13th AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment, AIIDE 2017

Y2 - 5 October 2017 through 9 October 2017

ER -

Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks

Abstract

Conference

Bibliographical note

Funding

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this