期刊论文

Entity Disambiguation Based on Wikipedia

中文

南华大学学报(自然科学版)

1673-0062

2016

30

2

83-86+92

10.3969/j.issn.1673-0062.2016.02.017

13B101:湖南省教育厅优秀青年项目 14YBA335:湖南省哲学社会科学基金 2015KG60:衡阳科学技术发展计划基金资助项目

本校为第一机构

实体排歧一直是信息抽取任务中的难点问题.传统的方法常常借助语言知识库来进行排歧,如WordNet等.但是这类方法存在着一个问题,就是很难对人名以及最新的词汇进行排歧.因此,本文利用当前流行的维基百科知识库中的排歧页对同名实体进行排歧处理,提出基于维基百科的实体排歧方法.实验显示该方法能够有效的对人名进行排歧处理并且效果显著.

Entity disambiguation has always been a difficult problem in information extrac-tion.The traditional method often uses the language knowledge-base to carry on the disam-biguation,such as Wordnet and so on.But there is a problem with this approach,it is diffi-cult to deal with name disambiguation. Therefore,in this paper,we use the disambiguation page of Wikipedia knowledge-base to deal with the same name,and put forward a method based on Wikipedia.Experiments show that the proposed method can effectively...