Information Extraction Based on Predicate-Argument Structures and FrameNet

• 作者: Feng Liu Jianliang Xu Le Kang
• 来源期刊: 世界华商经济年鉴·科技财经
• 关键词: Information Extraction Predicateargumentstructures Unsupervised FrameNet
• 年，卷（期）: 2012,（8）
• 页码范围: 102-103
• 语种: 中文

摘要

[Abstract]This paper presents a new method of unsupervised information extraction which takes advantage of predicate-argument structures and FrameNet[5]. The method makes a connection between predicate-argument structures and FrameNet. It presents an algorithm that makes one kind of arguments correspond to a slot of FrameNet. Evaluating on SemLink demonstrates that our method outperforms competitive unsupervised approaches by a wide margin.