On-Line Analytical Processing
On-Line Analytical Processing is a processing that supports the analysis of business trends and projections. It is also known as decision support processing and OLAP. An OLAP software enables companies to have real-time analysis of data stored in a database. An OLAP server is typically a separate component of an information system which contains specially coded algorithms and indexing tools to efficiently process data mining tasks with minimal impact on database performance.
OLAP uses multidimensional view of aggregate data to provide quick access to strategic information for further analysis and with this, a data user can have fast and very efficient view of data analysis as OLAP turns raw data into information that can be understood by users and manipulated in various ways. The multidimensional views of data that OLAP requires also come with packaged calculation-intensive capabilities and time intelligence.
OLAP is part of a wide category of business intelligence that includes ETL (extract transform load), relational reporting and data mining. Some critical areas of a business enterprise where OLAP are greatly used include business reporting for sales, marketing, management reporting, business process management (BPM), budgeting and forecasting, financial reporting and similar areas.
Those databases that are planned and configured for use with OLAP use a multidimensional data model which can enable complex analytical and ad-hoc queries with a rapid execution time. Outputs from an OLAP query are displayed in a matrix or pivot format with dimensions forming the row and column of the matrix; the measures, the values.
OLAP is a function of business intelligence software which can enable data end users to easily and selectively get extracted data and view them from different points of view. In many cases, an OLAP as aspects designed for managers who want to look to make sense of enterprise information as well as how the company fares well with the competition in a certain industry. OLAP tools structure data in a hierarchical manner which is exactly the way many business mangers thinks of their enterprises. But OLAP also allows business analysts to rotate data and change relationships and perspectives so they get deeper insights into corporate information to enable them to analyze historical as well as future trends and patterns.
The OLAP cube is found in the core of any OLAP system. The OLAP cube is also referred to as the multidimensional cube or a hypercube. This cube contains numeric facts which are called measures and these measures are further categorized into several dimensions. The metadata of the cube is often made from a snowflake schema or star schema of tables in a relational database. The hierarchy goes from measures which are derived from the records in the fact table and dimensions which are derived from dimension tables.
A claim has it that OLAP cubes for complex queries has the power to produce answers in around 0.1% of the time for the same query on OLTP relational data. Aggregation is the key for OLAP to achieve the amazing performance and in OLAP, aggregations are built from the fact table. This is done by changing the granularity on specific dimensions and aggregating up data along these dimensions.
There are different kinds of OLAP such as Multidimensional OLAP (MOLAP) which uses database structures which are optimal for attributes such as time period, location, product or account code; Relational OLAP (ROLAP) wherein the base data and the dimension tables are stored as relational tables and new table are created so they can hold aggregated information; and Hybrid OLAP (HOLAP) which can be a combination of OLAP types. Many software vendors have their own versions of OLAP implementations.