On-Line Analytical Processing
On-Line Analytical Processing
(database)OLAP primarily involves aggregating large amounts of diversedata. OLAP can involve millions of data items with complexrelationships. Its objective is to analyze theserelationships and look for patterns, trends, and exceptions.
The term was originally coined by Dr. Codd in 1993 with 12"rules". Since then, the OLAP Council, many vendors, andDr. Codd himself have added new requirements and confusion.
Richard Creeth and Nigel Pendse define OLAP as fast analysisof shared multidimensional information. Their definitionrequires the system to respond to users within about fiveseconds. It should support logical and statistical processingof results without the user having to program in a 4GL. Itshould implement all the security requirements forconfidentiality and concurrent update locking. The systemmust provide a multidimensional conceptual view of the data,including full support for multiple hierarchies. Otheraspects to consider include data duplication, RAM and diskspace requirements, performance, and integration with data warehouses.
Various bodies have attempted to come up with standards forOLAP, including The OLAP Council and the Analytical Solutions Forum (ASF), however, the Microsoft OLE DB for OLAP API is the most widely adopted and has become the de facto standard.
http://access.digex.net/~grimes/olap/.
Usenet newsgroup: news:comp.databases.olap.
http://arborsoft.com/papers/finkTOC.html.