Deep Web Interface Integration And Data Annotation Method

With the development of the Internet, the numbers of web grow quickly.there are a lot of information resources on the Internet, but because the Deep Web data is massive, heterogeneous, diversity, dynamic, the Features of Deep Web data make the data useless. The mainstream search engines can search static web on the Internet, but in fact, static web pages is only a small part of the web, most of the information on the Internet can't be searched, Deep Web is the part that cannot be indexed by traditional search engines, especially those generated by querying the online database, in recent years, Deep Web has attracted the attention of scholars, how to make the data resources hidden in the network useful which has became the hot spots. Integrated interface and data-tagging is an important thesis, this paper is started in this context.This paper undertake a study about Deep Web interface integration and data annotation, because there is not a integrated interface to obtain data and the result do not have semantic. The code of the HTML page is not irregular, in order to specific the code. This paper improves the tool and normative rules. In order to map the property between two interfaces, the paper use the method that user-match and table-match.The system solve the problem that property unmatched. About integrate interface, there is a method about Interface integration, which use pattern to integrate interface. About the query conversion, different interfaces have different methods. This paper designs a system to get the data that from different web. In the result-pattern-matched, in order to solve the problem that the result do not have semantic, the paper use the method that the match between interface and result,and the paper design a annotation system to solve the problem that the result data do not match and lack of semantic problems. The experiments achieve the desired results.