计算机工程2012，Vol.38Issue(1)：48-50,54,4.DOI:10.3969/j.issn.1000-3428.2012.01.012
基于子树特征的中文实体关系抽取
Chinese Entity Relation Extraction Based on Subtree Feature
姚全珠 1王美君 1李如琼1
作者信息
- 1. 西安理工大学计算机科学与工程学院,西安710048
- 折叠
摘要
Abstract
Kernel methods for relation have the implicit representation of feature spaces which can't distinguish useful feature from useless. As a result, it introduces noise and affect performance. Aiming at this problem, this paper presents entity relation extraction based on the feature of subtrees. The proposed method uses subtree mining and feature selection to get the more useful subtrees, and the feature vector is constructed on them for categorization. Experimental result in Chinese language database shows that the proposed method for entity relation extraction is effective.关键词
实体关系抽取/短语结构语法/依存语法/特征选择/卡方统计量Key words
entity relation extraction/ phrase structure grammar/ dependency grammar/ feature selection/ Chi-squared statistic分类
信息技术与安全科学引用本文复制引用
姚全珠,王美君,李如琼..基于子树特征的中文实体关系抽取[J].计算机工程,2012,38(1):48-50,54,4.
