期刊名称：International Journal of Computer and Information Technology
印刷版ISSN：2279-0764
出版年度：2015
卷号：4
期号：1
出版社：International Journal of Computer and Information Technology
摘要：In this paper, we have developed an approach to specify when updating materialized views in real-time data warehouse. The aim is to improve the maintenance load which may have a bad impact on the query response time especially in a real-time environment where the data updates are frequent and the freshness of served data is highly required. We have introduced a new concept “Interest center” which is a frequent set of user queries i.e. common and frequent user demands. We identify the interest centers from the historic of user queries. Then, periodically, we analyze the recent asked queries to identify the more probably interest centers of the current users. This analysis allows us to estimate the next future user queries which will be part of the identified interest centers. Consequently, we affect the online update to the materialized views that are sources for the estimated queries. For the rest of materialized views we apply the on-demand update. Our experiments showed that the proposed hybrid approach guarantees strong consistency of data and allows reducing latency of updating materialized views. In addition, they prove that our solution decreases the maintenance load significantly better that the on-demand and the online policies.
