Research And Implementation Of The Query Rewrite Method Of Materialized View In Information Integration

The goal of information integration system is building a global schema on theheterogeneous, autonomous, independent data sources. The global schema shields thedifference of the original data source and can provide users with a unified query interface.Sousers can ask a query request simply defined in the global schema and do not care how to getthe data.There are tow global schema in this pater’s information integration system:Intermediary view and Materialized view. Intermediary view is a virtual view and thequery request which based on it needs to be assigned to each data source to be executed.Materialized view can respond quickly to a query because of it has saved the query results whenthe view being created. At the same time, we need to update the data for the materialized view.In order to submit the incremental data to materialized view, we need to build the schema ofQuasi Underlying Table based on the different query capability of data sourses and rewrite thedefinition of the materialized view. In addition, we also need to rewrite the definition of thedistribution of target.In this paper, we research the problems of query rewrite process of materialized view basedon the Wrapper/Mediator architecture.The main problems which need to be resolved inquery rewrite process of materialized view are as follows:1. Query compilation：the lexical, grammatical, semantic analysis of the materialized viewdefinition;2. Query decomposition：How to decompose the definition to the single data source/singletable based on the different query capability of data sourses.3. Building the schema of Quasi Underlying Table: How to build the schema ofQuasi Underlying Table based on the different query capability of data sourses.4. DDL rewriting: How to decompose accurately and efficiently the definition ofmaterialized view to Quasi Underlying Tables and how to rewrite the definition of thedistribution of target.5. DQL rewriting: how to rewrite accurately and efficiently the query request based on theglobal schema.6. Query Optimization: the application of query optimization rules between query decomposition and query transformation.7. The internal representation of the executable file: convert the results of query rewritingto internal structure.