Shaping the Narrative Arc: An Information-Theoretic Approach to Collaborative Dialogue
arXiv:1901.11528
Abstract
We consider the problem of designing an artificial agent capable of interacting with humans in collaborative dialogue to produce creative, engaging narratives. In this task, the goal is to establish universe details, and to collaborate on an interesting story in that universe, through a series of natural dialogue exchanges. Our model can augment any probabilistic conversational agent by allowing it to reason about universe information established and what potential next utterances might reveal. Ideally, with each utterance, agents would reveal just enough information to add specificity and reduce ambiguity without limiting the conversation. We empirically show that our model allows control over the rate at which the agent reveals information and that doing so significantly improves accuracy in predicting the next line of dialogues from movies. We close with a case-study with four professional theatre performers, who preferred interactions with our model-augmented agent over an unaugmented agent.
arXiv comment20 pages, 9 figures
References in corpus (16)
- A Neural Conversational Model
- Universal Sentence Encoder
- Deep Reinforcement Learning for Dialogue Generation
- A Diversity-Promoting Objective Function for Neural Conversation Models
- From Once Upon a Time to Happily Ever After: Tracking Emotions in Novels and Fairy Tales
- A Persona-Based Neural Conversation Model
- Emotional Chatting Machine: Emotional Conversation Generation with Internal and External Memory
- A Neural Network Approach to Context-Sensitive Generation of Conversational Responses
- Recurrent Topic-Transition GAN for Visual Paragraph Generation
- Generating Wikipedia by Summarizing Long Sequences
- Learning End-to-End Goal-Oriented Dialog
- Generating Sentences by Editing Prototypes
- On Evaluating and Comparing Open Domain Dialog Systems
- Retrieve and Refine: Improved Sequence Generation Models For Dialogue
- Improvised Comedy as a Turing Test
- Predicting Movie Genres Based on Plot Summaries
Cited by in corpus (11)
- Latent Space Factorisation and Manipulation via Matrix Subspace Projection
- Logical Natural Language Generation from Open-Domain Tables
- Improved Natural Language Generation via Loss Truncation
- Data-to-Text Generation with Style Imitation
- Towards Faithfulness in Open Domain Table-to-text Generation from an Entity-centric View
- KGPT: Knowledge-Grounded Pre-Training for Data-to-Text Generation
- DYPLODOC: Dynamic Plots for Document Classification
- Exploring Versatile Generative Language Model Via Parameter-Efficient Transfer Learning
- Sentence-Level Content Planning and Style Specification for Neural Text Generation
- Performing Structured Improvisations with pre-trained Deep Learning Models
- Transforming Multi-Conditioned Generation from Meaning Representation