为了方便用户快捷高效的使用Deep Web中内容丰富、主题专一的高质量信息,对Deep Web数据源发现研究已成为一个非常迫切的问题。目前通用的方法是基于关键词的主题过滤策略,这样容易发现一些不相关的数据源,为此提出一种新的基于语义的Deep Web数据源聚焦爬行方法,利用朴素贝叶斯分类算法自动发现Deep Web数据源.实验验证了该方法的有效性。
To expediently utilize the rich ,oriented topic and high quality information of Deep Web, this problem on Deep Web data sources discovery has been focused by more and more people. Nowadays, topic filtering strategy based on key words is widely used, then it will obtain some irrelevant data sources. This paper proposes a new focused crawling method based on semantic for Deep Web data sources, and describes a technique for detecting query interface using naive Bayes classification. Finally, the method is validated by test.