Data Generalization is the process of creating successive layers of summary data in an evaluational database. It is a process of zooming out to get a broader view of a problem, trend or situation. It is also known as rolling-up data.

There are millions and millions of data stored in the database and this number continues to increase everyday as a company heads for growth. In fact, a group of process of process called extract, transform, load (ETL) is periodically performed in order to manage data within the data warehouse.

A data warehouse is a rich repository of data, most of which are historical data from a company. But in modern data warehouses, data could come from other sources. Having data from several sources greatly helps in the overall business intelligence system of a company. With diverse data sources, the company can have a broader perspective not just about the trends and pattern within the organization but of the global industrial trends are well.

In order to get a view of trends and patterns based on the analytical outputs of the business intelligence system can be a daunting task. With those millions of data, most of which disparate (but of course ironed out by the ETL process), it may be difficult to generate reports.

Dealing alone with big volumes of data for consistent delivery of business critical applications can already affect the network management tools of a company. Many companies have found that existing network management tools could hardly cope up with the great bulk of data required by the organization to monitor network and applications usage.

The existing tools could hardly capture, store and report on traffic with speed and granularity which are requirements for real network improvements. In order to keep the volume down to speed up network performance for effective delivery, some network tools discard the details. What they would do is convert some detailed data into hourly, daily or weekly summaries. This is the process called data generalization or as some database professionals call it, rolling up data. Ensuring network manageability is just one of the benefits of data generalization.

Data generalization can provide a great help in Online Analytical Processing (OLAP) technology. OLAP is used for providing quick answers to analytical queries which are by nature multidimensional. They are commonly used as part of a broader category of business intelligence. Since OLAP is used mostly for business reporting such as those for sales, marketing, management reporting, business process management and other related areas, having a better view of trends and patterns greatly speeds up these reports.

Data generalization is also especially beneficial in the implementation of an Online transaction processing (OLTP). OLTP refers to a class systems designed for managing and facilitating transaction oriented applications especially those involved with data entry and retrieval transaction processing. OLAP was created later than OLTP and had slight modifications from OLTP.

Many companies who have been using the relatively older OLTP cannot abandon OLTP’s requirements and re-engineer for OLAP. In order to "upgrade" OLTP to some degree, the information system department needs to create, manage and support a dual database system. The two databases are the operational database and the evaluational database. The operational database supplies data to be used to support OLTP.

The evaluational database on the other hand will supply data to be used to support OLAP. By creating these two databases, the company can be able to maximize the effectiveness of both OLAP and OLTP. The two databases will differ in the characteristics of data contained within and how the data is used. For instance, in the "currentness" attribute of data, the operational data is current while the evaluational data is historic.