The Research Of Data Search And Integration Technology Based On Materialized Views

The new millennium has seen a fast development of information technology and the expansion of information, and large scale databases of various types have been established one after another around the world. There also come emergent problems, such as how to make the best use of these databases, and how to integrate the distributed information of different databases. Researchers always have to face two challenges: 1) the realization of integrated searching and reduction of cost for information transportation online; 2) the enhancement of the efficiency in the process of implementation of integration of different structuring databases.In order to solve the problems of integrating distributed, different structuring databases, this project proposes an integrating scheme based on metadata and userschema. In this scheme the metadata of different databases are integrated into a metadata database with a uniform standard. The mapping between userschema and metadata are established to realize the information integration and search by using userschema concept and then deal with the problem of low efficiency to propose a method of materialized view to increase the search speed. The ultimate goal of this scheme is a solution to the common problem of data information integration through the related research, finally a united platform is established for data sharing and integration, and a virtual central database facing certain theme and central metadata is formed to support multi-area application.The finished work of this project includes the construction of a metadata database and the implementation of metadata and importing tools. Based on the previous work, the project's main research focuses are as following:1. How to support expert of certain area to build userschema on the basis of metadata, and create mapping between userschema and metadata to form a new searching concept for searching users;2. Supporting ultimate users to implement information searching based on userschema, including the actually information search from the source database based on the mapping informations, and presenting the final result with the userschema form through the linking and clearing of the intermediated results;3. Analyzing the efficiency of the search method, puting forward a optimization way by using materialized view, the optimization methods including the dynamic choosing algorithm of materialized view, matching algorithm between user's search expressions and materialized view and maintaining and updating algorithm of materialized view , and so on.4. Based on the research mentioned above, for this project,a search tool prototype system ( VOscherV1.0 ) has been designed and implemented for information integration.The system conducts the data search process with the optimization method of materialized view. Moreover, the usefulness and efficiency of the system has been proved by tests of downloading biology data from websites. With further improving and testing this sysem will be used in information integration system of human liver protein database.