摘要
随着数据库在Internet中的应用日益广泛,Deep Web集成(即Web数据库集成)成为当前信息领域的研究热点,模式匹配是Deep Web查询接口集成中的一个关键问题。目前大多数这方面的研究都是基于英文的,针对这种情况,探讨了中文Deep Web查询接口的模式匹配方法,并提出了一种基于《知网》、面向中文语义的模式匹配算法,并利用属性在查询接口上的相对位置信息解决语义冲突。手工收集查询表单对算法进行验证,实验表明该方法能使得接口之间属性匹配的正确率达到90%以上。
With the widespread application of database in the Internet , Deep Web integration （ that is, Web database integration） has become the hot research field of information, and schema matching is a key issue. At present, most research in this area are based on English, this article explored the Chinese Deep Web query interface schema matching methods. And pro- posed an approach based on HowNet for Chinese semantic schema matching algorithm and resolved the conflict in semantics making use of the relative location information of attributes in the query interface. The experiments show good performance of the schema matching approach as high accuracy as 90%.
出处
《计算机应用研究》
CSCD
北大核心
2009年第10期3750-3753,共4页
Application Research of Computers
关键词
深网
接口集成
模式匹配
Deep Web
interface integration
schema matching
