Materialized views, especially views involving aggregation, are often used in environments such as data warehouses to speed up the evaluation of complex queries. Because the views can be quite complex, as changes are made to the underlying base data it is usually better to incrementally maintain a view by propagating changes to base data onto the view than to recompute the view from scratch. Griffn and Libkin [GL95] provide a framework for deriving incremental view maintenance expressions for a view with duplicate semantics, where the view can be defined using any number of select, project, join, bag union, and monus bag-algebra operators. However, aggregation was considered only in a limited sense; for example, aggregation with groupby attributes was not allowed. We extend the framework of [GL95] to include maintaining views with aggregation in the general case, where aggregation can include groupby attributes and a view can include several aggregate operators
[1]
Gio Wiederhold,et al.
Incremental Recomputation of Active Relational Expressions
,
1991,
IEEE Trans. Knowl. Data Eng..
[2]
Venky Harinarayan,et al.
Implementing Data Cubes E ciently
,
1996
.
[3]
Alan R. Simon,et al.
Understanding the New SQL: A Complete Guide
,
1993
.
[4]
V. S. Subrahmanian,et al.
Maintaining views incrementally
,
1993,
SIGMOD Conference.
[5]
Per-Åke Larson,et al.
Eager Aggregation and Lazy Aggregation
,
1995,
VLDB.
[6]
Leonid Libkin,et al.
Incremental maintenance of views with duplicates
,
1995,
SIGMOD '95.
[7]
Hamid Pirahesh,et al.
Data Cube: A Relational Aggregation Operator Generalizing Group-By, Cross-Tab, and Sub-Totals
,
1996,
Proceedings of the Twelfth International Conference on Data Engineering.