الفهرس Only 14 pages are availabe for public view
 |  | from | 75 |  |  |
| | | from | 75 | | |
Abstract
Data warehouses collect and maintain a large amount of data iTom multiple distributed , autonomous information sources, in the form of materialized views in order to provide fast access to the integrated data. One of the major pr~blems in data warehousing is
materialized view maintenance. Although much attention has been given to this problem, there is a need for a new approach for single view structures that efficiently handles the occurrence of many updates at different information sources with a fewer number of join operations. There is also a need for a new approach that takes advantages of hierarchical view structures to perform a fewer number of join operation, to send a fewer number of messages and to guarantee a high level of consistency between materialized views and information sources.
This thesis introduces a new algorithm for incremental view maintenance in data warehouses. The proposed algorithm works for both single view structures and hierarchical view structures.
The ’proposed algorithm reduces the number of join operations performed and number of . messages sent to maintain all views in the structure while guaranteeing a strong consistency between materialized views and information sources. The proposed algorithm uses a divide and conquer strategy to reduce the number of join operations performed and hence to reduce the total maintenance time.
The maintenance workload in the proposed algorithm is distributed among all sites instead of being centralized at the source that contains the materialized view. Thus, the data warehouse is relieved iTom handling unnecessary intermediate messages before the view change is finally calculated.
Closed form solutions, for number of join operations performed and number of messages sent by the proposed algorithm, for the single view structure case, have been presented. Closed form solutions, for number of join operations performed and number of messages sent by the proposed algorithm, for hierarchical structures under some assumptions, have been also presented. A simulation study is conducted to compare the performance of the proposed algorithm to its counterparts, for both the single view structure case and the hierarchical view structure case.