| Summary: Data warehouses have played a very | | | | along dimensions. A three dimensional cube in this |
| important role in organizational settings in the recent | | | | example would have time, store and products as the |
| times. These can be used for sophisticated enterprise | | | | three dimensions. |
| intelligence systems that process queries required to | | | | Further, each dimension is divided into units called |
| discover trends and analyze critical factors in the | | | | members and the members of a dimension are |
| marketplace. These systems are known as online | | | | typically organized into a hierarchy. Similar members |
| analytical processing (OLAP) systems. OLAP systems | | | | are then grouped together as a level of the |
| help designers organize data in the warehouse | | | | hierarchy. For example, the top hierarchy level of a |
| distinctively. The data in data warehouses is | | | | time dimension can be years, with months at the |
| organized differently than in traditional transaction | | | | next level, then weeks, days and finally hours at the |
| processing databases. | | | | bottom level of the hierarchy. At each intersection of |
| OLAP systems are designed in an intention to handle | | | | the three dimensions, the values for the measures |
| the queries in an organization required to discover | | | | that match those three dimension values are |
| trends and critical factors. This type of queries | | | | recorded. |
| basically requires large amounts of data. OLAP data is | | | | When it comes to the specific dimensions and |
| always organized into multidimensional cubes. In other | | | | measures for the cubes in an OLAP system, the |
| words an OLAP structure created from the | | | | kinds of analysis come across as an important aspect. |
| operational data is called an OLAP cube. The cube is | | | | An OLAP system operates on OLAP data in data |
| created from a start schema of tables. In this type | | | | warehouses. The reason behind using OLAP in data |
| of schema, the fact table is placed at the center and | | | | warehousing is speed. OLAP systems provide rapid |
| linked to numerous dimension tables. The fact table | | | | access to large amounts of performance data from |
| contains the core facts, which make up the query. | | | | different viewpoints in order to assist business |
| Dimension tables indicate how the aggregations of | | | | analysts and managers throughout an enterprise. |
| relational data can be analyzed. | | | | There are three types of OLAP- Multidimensional |
| The multidimensional cube structure of data gives | | | | OLAP (MOLAP), Relational OLAP (ROLAP) and Hybrid |
| better performance for OLAP queries as compared | | | | OLAP (HOLAP), each with certain benefits. MOLAP |
| to the structure where data is organized in relational | | | | uses a summary database and creates the required |
| tables. The basic unit of a multidimensional cube is | | | | schema as a dimensional set of both base data and |
| called a measure. Measures are the units of data that | | | | aggregations. ROLAP utilizes relational databases. Here |
| are being analyzed. Take the example of a | | | | the base data and the dimension tables are stored as |
| corporation that operates hardware stores. Suppose | | | | relational tables and new tables are created to hold |
| it wants to analyze revenue and discounts for the | | | | the aggregation information. Hybrid OLAP uses |
| different products it sells. In this case, the measures | | | | relational tables to hold base data and |
| would be the number of units sold, revenue and the | | | | multi-dimensional tables to hold the speculative |
| sum of any discounts. These measures are organized | | | | aggregations. |