Typically, data flows from one or more online transaction processing (OLTP) databases into a data warehouse on a monthly, weekly, or daily basis.The data is normally processed in a staging file before being added to the data warehouse.I suggest that first we don't consider joins ..and only consider grouped aggregates over a single table. A fair portion of the community argues that everything that we do is merely materialized views :-) While materialized views are certainly related, there are however, things that are quite different in processing continuous queries over data streams.The first ideas on stream processing did come from work on scalable triggers which is intimately related to materialized views.For example, you can create a table to contain the sums of sales by region and by product.The summaries or aggregates that are referred to in this book and in literature on data warehousing are created in Oracle using a schema object called a materialized view.

Because spatial views are usually derived by complex spatial queries, it is desirable to actually materialize spatial views as opposed to being computed on demand for offering improved query performance. A main obstacle to spatial view materialization is that materialized views must be maintained in a consistent state with respect to the updates of source spatial objects. This paper defines the semantics of updating materialized spatial views in terms of the cardinality constraints of view derivation relationships between spatial view objects and source spatial objects. We can have a logging table that accumulates all the changes since the last refresh for this purpose.(This kind of table may exist anyway for replication or other purposes.) Finally, we could examine the changed rows, ignore the irrelevant ones, ignore the redundant ones (IE, rows that have been inserted and then deleted during that time), and decide whether doing a complete refresh would be quicker than doing several differential updates to the materialized view.