Telling Stories through Multi-User Dialogue by Modeling Character Relations

TL;DR

This paper presents a multi-task transformer model that improves character-driven story continuation by incorporating character relationships and personas, demonstrated on an extended Dungeons and Dragons dataset.

Contribution

It introduces a novel multi-task learning approach that integrates character relationships and personas into story continuation models, enhancing coherence and consistency.

Findings

Multi-task model outperforms baselines in story continuation accuracy.

Incorporating character relationships improves narrative coherence.

Extended dataset with automatically extracted relationships supports model training.

Abstract

This paper explores character-driven story continuation, in which the story emerges through characters' first- and second-person narration as well as dialogue -- requiring models to select language that is consistent with a character's persona and their relationships with other characters while following and advancing the story. We hypothesize that a multi-task model that trains on character dialogue plus character relationship information improves transformer-based story continuation. To this end, we extend the Critical Role Dungeons and Dragons Dataset (Rameshkumar and Bailey, 2020) -- consisting of dialogue transcripts of people collaboratively telling a story while playing the role-playing game Dungeons and Dragons -- with automatically extracted relationships between each pair of interacting characters as well as their personas. A series of ablations lend evidence to our hypothesis, showing that our multi-task model using character relationships improves story continuation accuracy over strong baselines.