首页> 外文会议>International conference on web information systems engineering >A Minwise Hashing Method for Addressing Relationship Extraction from Text
【24h】

A Minwise Hashing Method for Addressing Relationship Extraction from Text

机译:一种用于寻址文本关系提取的MeNIVES散列方法

获取原文

摘要

Relationship extraction concerns with the detection and classification of semantic relationships between entities mentioned in a collection of textual documents. This paper proposes a simple and on-line approach for addressing the automated extraction of semantic relations, based on the idea of nearest neighbor classification, and leveraging a minwise hashing method for measuring similarity between relationship instances. Experiments with three different datasets that are commonly used for benchmarking relationship extraction methods show promising results, both in terms of classification performance and scalability.
机译:关系提取问题与文本文档集合中提到的实体之间的语义关系的检测和分类。本文提出了一种简单而在线方法,用于解决语义关系的自动提取,基于最近的邻分类的思想,利用了一个致法的散列方法来测量关系实例之间的相似性。具有三种不同数据集的实验,通常用于基准关系提取方法显示出于分类性能和可扩展性方面的有希望的结果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号