The Research On Deep Web Interfaces Integration And Query Results Ranking

With the rapid development of World Wide Web,web information is on the emergence of the Internet more and more. And the progress of both network dynamic language and web database technology makes web information show the characteristics of dynamic and deepness. Furthermore,the amount of information is in a speedy growth.To present traditional search engines,in-depth web information,which is referred to as Depp Web,can not be indexed effectively.Information in Deep Web,which possesses the features of obvious structure,high value and vigorous growth,has attracted an increasing number of people to make a thorough study of it.Nevertheless,a lot of work is still at an exploratory stage,since Deep Web is simply an emerging field of research.This article will firstly explore and research Deep Web information search and relevant technology. Secondly, some key technologies are intensively studied and then corresponding approaches are put forward. Finally, the paper has raised a design structure of prototype system oriented to Deep Web vertical meta-search engine. The main research contents of this article include:(1) In the research respect of Deep Web data sources discovery, the article completes the search for Deep Web data sources through the design of focused crawler which is based on search engines. Moreover, a judgment method of query interfaces is brought forward in the meantime.(2) In the research respect of query interface integration, the paper proposes an interface schema matching method based on association rule and deals with complex pattern matching.(3) In the research respect of sorting query results, this paper presents a way of comprehensively calculating rankings for query results. From three aspects of quality from data sources, query relevance and query response feedback, it will figure out the evaluation scores for query results comprehensively, which is used to sort query results. (4) Based on analysis of present status and work principle of search engines, the article advances a design structure of system oriented to Deep Web vertical meta-search engine and makes a detailed analysis of critical modules.In this paper,related experiments have conducted to demonstrate that the methods and technology proposed are feasible and effective.