Comments on the general framework about Enterprise system Integration outlined by Hasselbring

The integration of systems which have been developed and evolved independently is one of today's major challenges in computer science. In a large spectrum of application areas the necessity of integrating pre-existing software systems is present and demands for applicable solutions. Problems of coupling and integrating heterogeneous database and information systems are for instance investigated in the database area. Building multi-database systems or federated database systems incorporating legacy systems is a big challenge. On the other hand, systems integration is an important challenge for the area of software engineering as well. Traditionally, the integration of heterogeneous information systems proceeds in a bottom-up process. Information stored in existing legacy systems is analyzed with respect to potential overlaps. The overlapping areas of related information sources are subsequently integrated. The integration of these isolated information systems is usually realized by means of mediators, federated database systems or such-like system architectures. Typical goals for the integration of existing information systems are the development of global applications that access the data from multiple sources as well as consistency management of information that is stored in related systems. From my research on the internet, I learnt that the process of integrating the heterogeneous information systems is being done through different viewpoints. I would like to share these viewpoints with the class..

• The process in which heterogeneous information systems are traditionally integrated in a bottom-up way, had some problems with this approach.
• The process in which heterogeneous information systems could ideally be integrated in a top-down way, to achieve more usable and scalable systems.
• A combined yo-yo approach to exploit the benefits of both strategies and to serve as a migration path from the traditional bottom-up approach towards an ideal top-down approach.

Semantic interoperability is necessary in this context to ensure that exchange of information makes sense - that the provider and requester of information have a common understanding of the "meaning" of the requested services and data.
Organizations face the challenging task of integrating their distributed organizational units, information systems, and business processes for improved operation and attainment of organizational goals. There is the difficulty of dealing with heterogeneous applications that use different formats (syntax) and apply different meanings (semantics) to the data. There is the difficulty of coordinating the workflow so as the disparate organizational units act as a harmonious whole. The broad scope of the enterprise integration problem precludes approaches that tackle the entire problem but rather requires approaches that address a limited but useful integration type. The various information integration types and how they are related to each other is poorly defined. The article presents an enterprise information integration framework that aims to coalesce parallel approaches towards integration so that the information integration problem can be better understood as a whole. The enterprise information integration framework defines four levels of the enterprise system to identify the obstacles and to define the information integration types encountered at each level. The framework is used to analyse the currently used technologies and promising technologies toward enterprise information integration. The analysis identifies general trends and gaps for further research.

Introduction

Integrating software applications is a top priority for IT directors and CIOs in today’s global enterprises. They recognize that well-integrated applications greatly enhance the productivity of employees. For example, a support representative can better assist the customer if he knows what problems he has experienced in the past, what products he has purchased, and where more-detailed information about those products can be found. However, all this information is probably stored in different software applications. For the support representative to jump from one application to another would be inefficient, especially with a customer waiting on the phone. Ideally, all this information would be easily accessible on a single computer screen.
Integrating the various applications would be fairly easy if they were all from same vendor or running on the same platform. Unfortunately, this is rarely the case. Today’s organizations are faced with the problem of integrating applications from different vendors running on different platforms. Mergers and acquisitions result in multiple disparate technologies and applications in use within a single enterprise. Even when a new application is purchased, it is impractical.

The integration challenge becomes even more complex when you consider the supply-chain activities of a retailer such as Wal-Mart, which has thousands of partners and suppliers. The inventory and ordering system at Wal-Mart must interface with thousands of supplier systems. Each of those suppliers work with hundreds of retailers. It is not likely that all involved systems use similar technologies or run on the same platform.

Web Services were created to address this challenge: the integration of heterogeneous technologies and applications. Web Services make it possible to glue together these disparate technologies within the enterprise or across different companies. In a sense, Web Services enable the creation of a “grid” of interoperating technologies and applications, which facilitate the “frictionless” modern commerce environment.