New Features for Relation Extraction

摘要：

The automatic extraction of relations between entities expressed in natural language text is an important problem for IR and text understanding. In this paper we show a new method for relations extraction. In the proposed method, effective PartialTrees were regarded as new features, and then vector was constructed on them for categorization. New features using is brought to avoid heuristic feature selection in relation extraction based on feature, at the same time the noise brought by kernel methods is filtered. The experimental results show that the proposed method for entity relation extraction is effective.

关键词： Relation Extraction PartialTrees New features Classification

作者: Yao Quan-Zhu Wang Mei-Jun Zhang Jiu-Long

作者单位: School of Computer Science & Engineering Xian University of Technology Xian, China

会议类型: 国际会议

会议名称: 2010 International Conference on Information Security and Artificial Intelligence(2010年信息安全与人工智能国际会议 ISAI 2010)

会议地点: 成都

会议语种:英文

页码: 399-403

在线出版日期: 2010-12-17（万方平台首次上网日期，不代表论文的发表时间）