Integration (software)

from Wikipedia, the free encyclopedia

The integration of computer science , especially in software engineering , serves to link different applications. In contrast to coupling , this involves reducing and avoiding interfaces . A distinction can be made between functional integration, data integration and business process integration.

Function integration

When integrating functions, several applications based on the division of labor are combined in one application. This avoids having to implement the same functions in different applications. At the same time, interfaces for data transfer are avoided.

Data integration

With data integration, all the applications involved access a uniform data model in order to avoid redundancies . The semantics are the same for all participating applications.

A disadvantage of these two forms of integration is that they can result in software that is more difficult to maintain . Function integration in particular can be problematic here. The simpler implementation is advantageous.

Business process integration

In business process integration, the various functions along the business processes are integrated with one another with the help of an integration platform. The individual applications that provide the functions remain unchanged, which also enables the use of standard software. This type of integration is a prerequisite for a successful and at the same time cost-effective implementation of e-business or business bus initiatives .

Software implementations

For the integration task in organizations, there are different approaches to software implementation in computer science. Some conceptual differences are apparently historical because over the years the overlap has become less and less due to the increase in functions. A typical characteristic of this situation is that many software products fall into several categories or even all categories. With many providers of software solutions it can be observed that these terms tend to be avoided. The conceptual differences between the following terms are still there today, but are difficult to recognize for newcomers, because with low requirements it can be implemented with all software categories. The differences are more in the economy. If a software solution supports more than one category well, then more flexible and more scalable solutions can be operated with a small team without having to override several different instances / systems during operation.