Page History

...

There are several ways to integrate data that depend on the size of the business, the need is fulfilled and the resources available.:

Manual data integration A process by which an individual user manually collects necessary data from various sources by accessing interfaces directly then cleans it up as needed and combines it into one warehouse. This is highly inefficient and inconsistent and makes little sense for all but the smallest of organizations with minimal data resources.

Middleware data integration is an integrated approach where a middleware application acts as a mediator, helping to normalize data and bring it into the master data pool. Middleware comes into play when a data integration system is unable to access data from one of these applications on its own.

Application-based integration is an approach to integration wherein software applications locate, retrieve, and integrate data. During integration, the software must make data from different systems compatible with one another so they can be transmitted from one source to another.

Uniform access integration is a type of data integration that focuses on creating a front end that makes data appear consistent when accessed from different sources. The data, however, is left within the original source. Using this method, object-oriented database management systems can be used to create the appearance of uniformity between unlike databases.

Common storage integration is the most frequently used approach to storage within data integration. A copy of data from the original source is kept in the integrated system and processed for a unified view. This is opposed to uniform access, which leaves data in the source. The common storage approach is the underlying principle behind the traditional data warehousing solution.

...