Does Linguistic Structure Enrichment Enhance Coherence Assessment? Not With Current Architectures

2026-09-09

AI Summary

研究探讨了通过增加语法和修辞信息是否能改善文本不连贯性预测，但发现纯文本因与当前语言模型架构更兼容而表现更好。

Abstract

Recent advances in large language models have transformed human-computer interaction. Despite their fluency, these models often produce texts that are grammatically correct but semantically incoherent, containing contradictions or disruptions in logical flow. This work investigates whether enriching text with syntactic and rhetorical information can improve incoherence prediction. Our experiments and analysis show that plain texts achieved higher accuracy because the added information was structurally and syntactically incompatible with the language model's architecture. Additionally, to demonstrate the practical importance of coherence assessment, we performed zero-shot experiments on a Brazilian disinformation dataset, suggesting that textual coherence can serve as a proxy for detecting misleading content. Code and models are available at https://github.com/ittozzamV/cohereclassifier.

Problem

• incoherence
• large language models
• syntactic and rhetorical information
• textual coherence

Innovation

• syntactic and rhetorical information
• incoherence prediction
• language model's architecture
• zero-shot experiments
• disinformation