A mechanism for view consistency in a data warehousing system

Abstract

A data warehouse is a repository of integrated information from distributed, autonomous and heterogeneous information sources. Materialized views in the data warehouse must be maintained when updating the data in the information sources. The conventionally used incremental view maintenance algorithm causes the problem of anomalies. In addition, previous attempts to solve anomalies focus on compensating the changes of the view. In this work, a novel method which uses the information already available at the data warehouse instead of performing the compensation is presented. Rules are used to predict information deemed necessary to maintain a materialized view. The information is stored as auxiliary views. The proposed method does not require any data transmission between the information sources and warehouse when a change of a materialized view from an updated message occurs in information sources. Moreover, the method proposed herein saves a significant amount of time when materialized views are incrementally maintained.