Skeleton-based Coherence Modeling in Narratives

arXiv cs.AI / 4/6/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper investigates whether the consistency of extracted “sentence skeletons” across adjacent sentences is a reliable metric for evaluating narrative coherence.
It introduces a Sentence/Skeleton Similarity Network (SSN) to model coherence from sentence/skeleton pairs and reports that SSN outperforms simple baseline similarity measures like cosine similarity and Euclidean distance.
Despite the promise of skeleton representations, the authors find that sentence-level coherence models still outperform skeleton-based approaches for coherence evaluation.
The results suggest current NLP coherence modeling progress is aligned with using full sentences rather than relying primarily on sub-components like skeletons.

Abstract

Modeling coherence in text has been a task that has excited NLP researchers since a long time. It has applications in detecting incoherent structures and helping the author fix them. There has been recent work in using neural networks to extract a skeleton from one sentence, and then use that skeleton to generate the next sentence for coherent narrative story generation. In this project, we aim to study if the consistency of skeletons across subsequent sentences is a good metric to characterize the coherence of a given body of text. We propose a new Sentence/Skeleton Similarity Network (SSN) for modeling coherence across pairs of sentences, and show that this network performs much better than baseline similarity techniques like cosine similarity and Euclidean distance. Although skeletons appear to be promising candidates for modeling coherence, our results show that sentence-level models outperform those on skeletons for evaluating textual coherence, thus indicating that the current state-of-the-art coherence modeling techniques are going in the right direction by dealing with sentences rather than their sub-parts.