汉语部分实体的抽取模型

An Extraction Model of Recognizing Chinese Partial Entity

为了识别汉语实体 ,提出了利用模式匹配技术构造一个抽取模型 .为了识别该模型的有效性 ,系统的测试语料采用 Internet网上真实语料 ,语料全部来自新加坡联合早报 ,主要是新闻领域 ,经过开放测试 ,该系统识别实体召回率为 46 .0 2 % ,准确率为 5 2 .94% .

In order to recognize the Chinese Entity,the paper introduces an extraction model based on pattern matching technique.All test corpora come from internet and most of them are news corpora.Initial running of the system on a test corpus produced the following result the recall rate of 46 02% and the accurate rate of 52 94%.