中文摘要：

Deep Web中蕴含了海量可供访问的信息,如何构建一个具有较好适用性和高效数据处理能力的Deep Web数据集成系统是有效利用Deep Web信息的关键.提出一种基于结果模式的DeepWeb数据集成机制,通过结果模式可以实现高效的数据抽取,并且在结果模式的基础上可以根据用户查询请求动态生成结果输出模式,为高效的查询结果处理奠定了良好基础;同时,针对Deep Web数据源特点,给出数据源间冲突的分类及解决策略,为解决数据源间的异构问题奠定了良好基础.

英文摘要：

With the rapid development of Deep Web,how to construct a Deep Web data integration system with good applicability and high ability of data processing becomes more and more important. The paper proposes the mechanism of Deep Web data integration based on result pattern. Through result pattern highly efficient Deep Web data extraction can be realized. According to the users＇ query the result output pattern can be dynamic generated on the basis of result pattern,and lays the well foundation for efficient query result processing. At the same time,the paper gives the classification and solution strategy of conflicts among Deep Web data sources that lay the well foundation for resolving the heterogeneity among Deep Web data sources.