研究人员开发了一种名为全息神经PCFG（Hol-PCFG）的新方法，用于无监督成分句法分析。该方法利用语法符号嵌入之间的代数关系来模拟PCFG规则评分，并改编全息嵌入来表示左子节点、右子节点和词汇发射等关系。Hol-PCFG在六种语言上取得了最先进的成果，显著减少了规则评分参数并提高了训练稳定性。值得注意的是，它可以在没有预先形态学分段的情况下直接从字符解析日语，同时保持高性能。 AI
影响 这种新的解析方法提供了更高的效率和可解释性，有望推进自然语言理解能力。
排序理由 该集群包含一篇详细介绍无监督解析新方法的学术论文。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
arXiv:2607.08063v1 Announce Type: new Abstract: Unsupervised constituency parsing aims to accurately induce latent tree structures from raw text alone. Recent neural parameterizations of PCFGs achieve strong performance in both supervised and unsupervised parsing, yet rely on hig…
Unsupervised constituency parsing aims to accurately induce latent tree structures from raw text alone. Recent neural parameterizations of PCFGs achieve strong performance in both supervised and unsupervised parsing, yet rely on high-capacity black-box networks for rule scoring -…
Unsupervised constituency parsing aims to accurately induce latent tree structures from raw text alone. Recent neural parameterizations of PCFGs achieve strong performance in both supervised and unsupervised parsing, yet rely on high-capacity black-box networks for rule scoring -…