Research On Deep Web Data Integration And Application On Book Purchase Field

Owning to that the query interface forms must be filled in by users to access backend databases for acquiring the information of Deep Web, it's a problem that traditional search engines is unable to provide indexing service in the face of the current emerging Deep Web which contain richer information.Working principle, system structure and related technology of Deep Web data integration system has been analyzed for standing around existing problem, firstly. Based on that, the chief modules and related algorithm of the Deep Web data integrating system have been researched and discussed. Properties of this paper are mainly about:1. A novel interface-query pattern extraction based on web pages marking tree models has been advanced, where the object model is used for expressing query interface. Also, the specific realizing algorithm is given.2. After researching the algorithm of semblance among elements, the WDB(web data base) chosen function has been realized via calculating the relevance between the interface-query form element and subject domain for selecting data source.3. With the deep study of modeling done on the query conversion mechanism, the solution is provided to achieve mapping between integrated query interface and WDB query interface. Thereby, efficiency of query conversion has been improved.Based on the research of Deep Web data integration, design a Deep Web data integration system on book purchase area in order to obtain information from Deep Web automatically, accurately and effectively. Ultimately, the feasibility of this method has been proved by experimental analysis.