Story Realization: Expanding Plot Events into Sentences

TL;DR

This paper introduces an ensemble-based neural model that expands plot events into coherent sentences, improving automated story generation by maintaining semantic relevance and coherence, validated through human studies.

Contribution

The paper presents a novel ensemble approach for event-to-sentence generation that enhances story coherence and plausibility in neural story generation systems.

Findings

Generated stories are more coherent and plausible.

Human evaluations favor the proposed method over baselines.

The approach effectively preserves event semantics during sentence realization.

Abstract

Neural network based approaches to automated story plot generation attempt to learn how to generate novel plots from a corpus of natural language plot summaries. Prior work has shown that a semantic abstraction of sentences called events improves neural plot generation and and allows one to decompose the problem into: (1) the generation of a sequence of events (event-to-event) and (2) the transformation of these events into natural language sentences (event-to-sentence). However, typical neural language generation approaches to event-to-sentence can ignore the event details and produce grammatically-correct but semantically-unrelated sentences. We present an ensemble-based model that generates natural language guided by events.We provide results---including a human subjects study---for a full end-to-end automated story generation system showing that our method generates more coherent and plausible stories than baseline approaches.