The Research Of Optimizing Queries Based On Materialized Views

With the advent of the World Wide Web, a large number of information sources are available to users. It is more important to make commercial strategy from large amounts of data. Traditionally, the database system can not satisfy the user's need, and then introduce the data warehouse technology. Data warehouse technology introduces the concept of materialized view. Through storing the data that the queries need before quering, materialized views have improved the speed of quering greatly. Based on materialized views, the paper realizes query optimization through query rewritings.The paper introduces and analyses the actuality of materialized view technology. The materialized views selection, maintenance, and query are correlated closely. With lots of materialized views in the data warehouse, users can quickly get answering of queries by using materialized views. In data warehouse you should choose appropriate views to materialize, which can reduce the total response time and reduce the cost of views maintenance.When rewriting conjunctive queries using views, we present the bucket improved algorithm, which is based on the bucket algorithm. Since the algorithm has been presented the bucket enhanced, it can generate only sound solutions and also eliminate the need for an exponential conjunctive query containment tests. Especially as the number of sources is large, the algorithm enables such a system to rewrite queries efficiently, reduces the response time to users, improves its scalability and makes that is feasible for a system to handle large number of information sources. It provides proofs of the soundness and completeness for the algorithm.When both queries and views with built-in predicates, it becomes general conjunctive query. We present the algorithm based on goals for rewriting general conjunctive queries using views, it can response to users more quickly. And it tends to be more efficient when the constraints of the query involve only distinguished variables. It can greatly reduce the response to query, and optimize the queries.