Generating Narrative Text in a Switching Dynamical System

Early work on narrative modeling used explicit plans and goals to generate stories, but the language generation itself was restricted and inflexible. Modern methods use language models for more robust generation, but often lack an explicit representation of the scaffolding and dynamics that guide a coherent narrative. This paper introduces a new model that integrates explicit narrative structure with neural language models, formalizing narrative modeling as a Switching Linear Dynamical System (SLDS). A SLDS is a dynamical system in which the latent dynamics of the system (i.e. how the state vector transforms over time) is controlled by top-level discrete switching variables. The switching variables represent narrative structure (e.g., sentiment or discourse states), while the latent state vector encodes information on the current state of the narrative. This probabilistic formulation allows us to control generation, and can be learned in a semi-supervised fashion using both labeled and unlabeled data. Additionally, we derive a Gibbs sampler for our model that can fill in arbitrary parts of the narrative, guided by the switching variables. Our filled-in (English language) narratives outperform several baselines on both automatic and human evaluations.

[1]  Hugh Robert Macdonald,et al.  Palm Sunday , 1968 .

[2]  Dongyan Zhao,et al.  Plan-And-Write: Towards Better Automatic Storytelling , 2018, AAAI.

[3]  Nathanael Chambers,et al.  A Corpus and Cloze Evaluation for Deeper Understanding of Commonsense Stories , 2016, NAACL.

[4]  Mark O. Riedl,et al.  Event Representations for Automated Story Generation with Deep Neural Nets , 2017, AAAI.

[5]  Dina Sherzer,et al.  A dictionary of narratology , 1989 .

[6]  Kevin Murphy,et al.  Switching Kalman Filters , 1998 .

[7]  Yann Dauphin,et al.  Hierarchical Neural Story Generation , 2018, ACL.

[8]  Robert Michael Young,et al.  Story planning as exploratory creativity: Techniques for expanding the narrative search space , 2006, New Generation Computing.

[9]  S. Turner Minstrel: a computer model of creativity and storytelling , 1993 .

[10]  James R. Meehan,et al.  TALE-SPIN, An Interactive Program that Writes Stories , 1977, IJCAI.

[11]  Chris Donahue,et al.  Enabling Language Models to Fill in the Blanks , 2020, ACL.

[12]  Chin-Yew Lin,et al.  ROUGE: A Package for Automatic Evaluation of Summaries , 2004, ACL 2004.

[13]  Ole Winther,et al.  Sequential Neural Models with Stochastic Layers , 2016, NIPS.

[14]  Uri Shalit,et al.  Deep Kalman Filters , 2015, ArXiv.

[15]  Alex Wang,et al.  BERT has a Mouth, and It Must Speak: BERT as a Markov Random Field Language Model , 2019, Proceedings of the Workshop on Methods for Optimizing and Evaluating Neural Language Generation.

[16]  Christopher M. Danforth,et al.  The emotional arcs of stories are dominated by six basic shapes , 2016, EPJ Data Science.

[17]  Vladimir Propp,et al.  Morphology of the folktale , 1959 .

[18]  Ole Winther,et al.  A Disentangled Recognition and Nonlinear Dynamics Model for Unsupervised Learning , 2017, NIPS.

[19]  Max Welling,et al.  Auto-Encoding Variational Bayes , 2013, ICLR.

[20]  Dan Roth,et al.  Story Comprehension for Predicting What Happens Next , 2017, EMNLP.

[21]  Robert Michael Young,et al.  Narrative Planning: Balancing Plot and Character , 2010, J. Artif. Intell. Res..

[22]  K. Ito,et al.  On State Estimation in Switching Environments , 1970 .

[23]  Uri Shalit,et al.  Structured Inference Networks for Nonlinear State Space Models , 2016, AAAI.

[24]  Max Welling,et al.  Semi-supervised Learning with Deep Generative Models , 2014, NIPS.

[25]  David Barber,et al.  Expectation Correction for Smoothed Inference in Switching Linear Dynamical Systems , 2006, J. Mach. Learn. Res..

[26]  Scott W. Linderman,et al.  Recurrent switching linear dynamical systems , 2016, 1610.08466.

[27]  M. Athans,et al.  State Estimation for Discrete Systems with Switching Parameters , 1978, IEEE Transactions on Aerospace and Electronic Systems.

[28]  Eric Gilbert,et al.  VADER: A Parsimonious Rule-Based Model for Sentiment Analysis of Social Media Text , 2014, ICWSM.

[29]  Maximilian Karl,et al.  Deep Variational Bayes Filters: Unsupervised Learning of State Space Models from Raw Data , 2016, ICLR.

[30]  Andrew S. Gordon,et al.  Creative Help: A Story Writing Assistant , 2015, ICIDS.

[31]  Yann Dauphin,et al.  Strategies for Structuring Story Generation , 2019, ACL.

[32]  Timothy Baldwin,et al.  Narrative Modeling with Memory Chains and Semantic Supervision , 2018, ACL.

[33]  Mark O. Riedl,et al.  Controllable Neural Story Generation via Reinforcement Learning , 2018, ArXiv.

[34]  Mark O. Riedl,et al.  Toward Automated Story Generation with Markov Chain Monte Carlo Methods and Deep Neural Networks , 2021, AIIDE Workshops.

[35]  Michael Lebowitz,et al.  Story-telling as planning and learning , 1985 .

[36]  Ben Poole,et al.  Categorical Reparameterization with Gumbel-Softmax , 2016, ICLR.

[37]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[38]  P. Thorndyke Cognitive structures in comprehension and memory of narrative discourse , 1977, Cognitive Psychology.

[39]  Alon Lavie,et al.  METEOR: An Automatic Metric for MT Evaluation with Improved Correlation with Human Judgments , 2005, IEEvaluation@ACL.

[40]  Mark O. Riedl,et al.  Controllable Neural Story Plot Generation via Reward Shaping , 2018, IJCAI.

[41]  Xu Sun,et al.  A Skeleton-Based Model for Promoting Coherence Among Sentences in Narrative Story Generation , 2018, EMNLP.

[42]  G. Lakoff Structural Complexity in Fairy Tales , 1972 .

[43]  Alexander M. Rush,et al.  Learning Neural Templates for Text Generation , 2018, EMNLP.

[44]  Julian Togelius,et al.  DeepTingle , 2017, ICCC.