Service-oriented Data Integration Framework

With the fast development of the information technology, Web has already accumulated vast amounts of information and increasing very rapidly.In order to make use of the large number of valuable informations, enterprises and individuals are search for the effective technologies and methods to mine and fuse the useful information from the Web. However, as a result of the Web is very hugeness and complex, these are still so many difficult problems wait to be sloved in the Web service oriented data integration domain.At present, the research of Web services oriented data integration research focused on Deep Web research.Its main difficulties are focus on Web services discover and Web service combination research domain, mainly including the recall rate and precision rate in Web service discovery is not high, and Web service composition is too simple,low efficiency and low satisfaction.In this paper, proposed a kind of service-oriented data integration framework base on the improving of two key technologies including Web service discovery and Web service composition, and provide a heterogeneous environment Web services integration platform, ensuring the efficient Web service discovery and services compositon results of a high success rate and high satisfactionThis paper's search of Web service discovery: This paper use ontology concepts to describe the Web services, enabling the machine can discover the relations between Web services by ontology logic deduction, making system understand user's request semantically. This paper presents Web service interface discovery mechanism which base on traditional search engine and query conversion;presents an effective Web service discovery mechanism.The mechanism can work both statically and dynamically, that is using a static way to save the Web service interface which have already be found by the system, if system can't find the related Web services in the Web service repository, the system will dynamically invoke the Web service interface crawler to find the Web services and save it into the Web service library. The mechanism will run Web service interface crawler in the background mode to get new Web service interfaces and updates the Web service library continuously. That makes system discover Web service more effectively and adequately.This paper's search of Web service composition: this paper presents ontology-based cost graph algorithm for Web service composition.The algorithm integrate several factor include the execution cost,the user's orientation and service quality through a cost calculate formula, making the results not only satisfy service quality requirement but also is meet user's tendency. The system define Web service combination cost graph which is a directed acyclic graph and take Web service as node, take the relationships between Web services as edge, take the flow cost as the weight of edge. The system invokes the service-based depth-first search algorithm in the Web service combination cost graph, and find combination strategy.And then write the path information and combination strategy into the selected Web services by using this paper's combination strategy storage method. Next, Analysis and implemente the combination strategy and get the result service.Finally, according to the previous chapters of the Web service discovery and the Web service composition method base on ontology cost graph, devised a service-oriented data integration framework, and described the main function modules in detail, and designed a simulation test platform base on the framework.Through a test case, introduced the overall operation of the system. First, ontology map the user's requirement to relevant candidate Web services set, and then run web service composition base on ontology cost graph algorithm to calculate a combination strategy. The experiment indicated that this service combinationg strategy's implementation results can satisfie user's demand very well, and proven the correctness of service composition algorithm, also proven that this paper's service-oriented data integration framework is a effective mechanism to accurately identify user's requirements, reasonably discover and integrate related web services.