中文Web概念挖掘系统设计与测评

Design & Test of the Web Concept Mining System in Chinese

Web概念挖掘系统的总体设计思想是 :基于《中图法》知识库的主题标引和自动分类 ,即依据标引源权重方案进行文本的概念提取 ,利用语义相似度算法进行文本的自动分类 .本文简要介绍了概念挖掘系统的基本情况 ,并进行了 4种加权标引方案的比较和性能的测评 .

The paper introduced the web concept mining system in Chinese, tested and evaluated the efficiency & function of the system. Based on the automatic indexing function of the Chinese Web Concept Mining System, we sample 150 web pages of economic science at random, index them manually and assign them the class numbers automatically. The statistics are worked out on the conforming between the manual indexing and the automatic indexing in order to compare four indexing schemes to prove the feasiblity of the Web Concept Mining System.