Towards Controllable Story Generation

Abstract

We present a general framework of analyzing existing story corpora to generate controllable and creative new stories. The proposed framework needs little manual annotation to achieve controllable story generation. It creates a new interface for humans to interact with computers to generate personalized stories. We apply the framework to build recurrent neural network (RNN)-based generation models to control story ending valence 1 (Egidi and Gerrig, 2009) and storyline. Experiments show that our methods successfully achieve the control and enhance the coherence of stories through introducing storylines. with additional control factors, the generation model gets lower perplexity, and yields more coherent stories that are faithful to the control factors according to human evaluation.

Unsupervised Hierarchical Story Infilling

“…Recently, the field has been influenced by the success of (conditional) neural language models (Bengio et al, 2003;Schwenk and Gauvain, 2004;Bahdanau et al, 2015;Nallapati et al, 2016). Story generation with neural models include (Chourdakis and Reiss, 2017; Peng et al, 2018;Radford et al, 2019).…”

A Hybrid Model for Globally Coherent Story Generation

“…Noting that RNN based language models could only account for local coherence, attempts have been made to perform text planning on a higher level (e.g. Jain et al, 2017;Peng et al, 2018). Puduppully et al (2018) performs content selection and planning with attention based neural networks before surface realization, to generate specifically structured NBA game summaries.…”

Proceedings of the First Workshop on Narrative Understanding

“…Recently, the field has been influenced by the success of (conditional) neural language models (Bengio et al, 2003;Schwenk and Gauvain, 2004;Bahdanau et al, 2015;Nallapati et al, 2016). Story generation with neural models include (Chourdakis and Reiss, 2017;Peng et al, 2018;.…”

“…This evolution has sparked the development of tools to assist human writers. For instance, Fan et al (2018b) suggest generating short stories from high-level prompts, Clark et al (2018b) study the interaction of human and language models for creative writing, and Peng et al (2018) propose an interactive control of story lines. In addition, products such as Grammarly offer suggestions to improve grammar and wording (Hoover et al, 2015).…”