The Deepweb Interface Schema Matching Method Based On Correlation Mining

Along with technical unceasing development, people have already been unfamiliar with Internet and network resources. And with the time advancement, the data resources of the network has been increasing, these astonishing data resources formed a number of online databases to hide in the deep web. Users can access a variety of online databases and find related resources according to their own need. The most direct way is to fill a query form in the web page, and to access the online databases through the query interface. However, Different query interface contains the different attribute information, and there are heterogeneous situations in the database schema. In other words, there are heterogeneous situations in the query interface of different online database, although the synonymous words cannot carry on the direct inquiry. Moreover, as the various domains information constantly expanding in the web, Users want to find some information which has high recall and high precision has become a difficult problem. If we can find the sequence which has the matched of attributes in the different query interface, User input inquiry information through different query interface can be mapped to this uniform sequence. And this method will avoid search different web pages in order to inquire some information. Therefore, to solve the above problems, this article will do the schema matching on the different quire interface.In the needs of DeepWeb query interface schema matching research, based on the result of research of the DeepWeb query interface schema matching both in domestic and abroad, the paper provide an effective schema matching method which base on correlation mining. In the module of the attribute matching, this paper proposed a Simple and highly effective correlation algorithm to ensure the accuracy of matching. Experiments proved that the whole process is a low complexity and has the high accuracy rate.