Online analytical processing (OLAP) is a method of using multidimensional databases to support quick reporting, frequently involving trend-analysis. The primary query language for OLAP is called Multidimensional Expressions (MDX). Its name is derived from the program class known as online transactional processing (OLTP). Online analytical processing is a technique of data analysis used in the business intelligence (BI) field.
BI involves using technology to analyze an organization’s internal processes and data to support its decision making. When using online analytical processing for BI, historic data is often the subject of the analysis, but BI can also encompass analysis of current and future states. Along with OLAP, other data management techniques that fall into the realm of BI include data mining, reporting, operational performance management, and predictive analytics.
Online analytical processing is frequently used for ad hoc reporting, and typically generates reports in a pivot or matrix format. Departments that may make use of OLAP include finance, operations, sales, and marketing. Types of uses can include budgeting and forecasting.
One of the defining characteristics of online analytical processing is the OLAP cube. The concept of the cube correlates the elements known as measures and dimensions, which describe the various measures’ metadata. A relational database’s snowflake or star schema tables may be the source of the metadata. An example of a cube is using a business’ individual accounts receivable amount as a measure, with a due date as a dimension.
OLAP uses databases that are designed with multiple dimensions. These databases may be smaller than those needed for the data warehousing capabilities that are often used for business intelligence. Compared to other types of analysis, fewer details of transaction are usually needed in online analytical processing. Not only are the OLAP databases often smaller than data warehouses, accessing the OLAP databases is often faster than accessing relational databases.
There are various specialties of online transaction processing. Several of the more frequently used specialities include multidimensional, relational, and hybrid. Multidimensional OLAP stores data in multidimensional arrays, relational OLAP uses relational databases, and hybrid OLAP uses a combination of the relational and specialized tables.
Though online transactional processing is an important technique in BI, more sophisticated tools or improvements to OLAP may be required for organizations that are interested in predictive analysis and business analytics. Predictive analysis is frequently used to forecast events such as customer buying behavior. Business performance data is usually the target of business analytics.
