【摘要】 论文论述了Web文档的信息抽取的基本方法,设计并实现了一个基于Internet上的军事演习信息抽取系统—SBIES。在系统中引入了分装器的机器学习算法来获取网页抽取规则,采用基于最大熵模型的组块分析方法进行部分语法分析,利用模式匹配的方法实现信息的自动抽取,以数据库与XML相结合的方式组织信息库,并实现信息的Web表示和查询。系统测试结果表明,它具有较高的抽取召回率和抽准确率。更多还原

【Abstract】 Information Extraction plays an important role in knowledge acquisition and information service.This paper discusses briefly the key techniques for information extraction,and it designs and implements a Sham Battle Information Extraction System(SBIES).It constructs automatically wrappers by machine learning algorithms,applies Maximum Entropy model to conduct Chinese chunk parsing and makes use of a sets of extraction patterns to extract specific information and relationships from relevant HTML documents.Moreover,it also combines the XML expression with the organization of database,so it realizes the presentation and query of information extracted based on Web.It shows higher recall and precision by testing SBIES.更多还原