Research On Web Resource Management And Web Service Performance

Along with the rapid development of Internet techniques and the universal popularization of Internet applications, the number of Web resources is increasing exponentially. Though there exist hundreds of Web Searching Engines, Personalization Web Information Recommendation Systems and Web Query Languages, the management of Web resources and the qualities of Web services are getting worse due to its characteristics of distribution, opening and isomer. This dissertation is mainly about the Technique of Web Database Management. First, it points out the disadvantages of the current methods in Web Database Management on data model, resource design, heterogeneous resource integrating, browsing, information retrieval and reconstruction. Then it presents a new description model of Web resource and discusses its application in Web Database Management. With the new data model and a series of improvements in techniques, we can improve the performance of Web Database Management wholly. The main contributions and innovations of this dissertation are as follows: (1) To overcome the disadvantages of the traditional description method of Web resource concerning resource design, data redundancy and consistency, a virtual description model of Web resource is presented. The new model not only conveniently supports information share at data level and can make the resource renewed dynamically with the change of its content, but also can be regarded as a common data model for the integration of heterogeneous Web resource. (2) Based on the new model, a modularization design method of Web resource (HTML and XML resources) has been given by introducing the concepts about the module node, the grafting operation and pruning operation of ETG. The new technique is capable of information share at design level, and allows the designer to design individually a static and dynamic common information module that can be used to build a Web resource. It has the advantages of simplifying the design of Web resource, reducing the design redundancy, and it is easy to fast update the resource design and to maintain its consistency. In general, the effect of the new technique in the design of Web resource is the same as the modularization design method in the design of program. (3) Based on the new model, an architecture for the integration of heterogeneous Web resources has been given by analyzing the various current integrating techniques on heterogeneous data resources and application systems. In the design of wrappers, we combine the data warehouse method and the virtual method. Compared with the integration systems using the data warehouse method, the new system has the advantages of reducing the data redundancy, being easy to maintain data consistency. Compared with the integration systems using the virtual method, the new system has the following advantages: it need not resolve an integration query to the sub-queries; it need not translate them to the queries of different application systems; it can simplify the design of the wrapper; it is easy to add a new heterogeneous resource. (4) Based on the new model, a new technique of extracting information from Web resource and compounding them into a new Web resource has been given by introducing the concepts of ETG operations and ETG reconstruction. A tag query language (TagSQL) has also been given as a user interface. It has many advantages such as the powerful operating functions, the convenience of its use, the flexibility of extracting information, the capability of compounding new Web resource based on Web resources. It is a complement for the Web information searching system that searches Web resources based on their contents. The new technique can be extended and applied in our integrated system of heterogeneous Web resources. (5) Based on the new model, a fast browsing technique of Web resource has been presented by introducing a content module about the Web resource. It can speed up the process of browsing Web resource by avoiding the transmission of uninterested Web resources or un...