Component Based Approach to Building Data Integration Tools

Abstract

A method facilitating interoperability across the integrating platforms is provided by composing purpose specific integrating tools using conceptual modeling framework and a set of primitive foundational components. The model driven architecture provides easy migration from one solution architecture to another. Also, the present invention discloses a method to migrate from one implementation to another using set of composed purpose specific integration tools.

Description

FIELD OF THE INVENTION[0001]The present invention relates to data integration and more particularly, to a method of composing purpose specific integration tools for facilitating migration from one integrating platform solution to another.BACKGROUND OF THE INVENTION[0002]The term data integration refers to the problem of combining data residing in heterogeneous sources and currently, it relates to wide range of technologies, from extract, transform and load (ETL) to enterprise application integration (EAI) to enterprise information integration (EII) and various change propagation technologies. There has been extensive theoretical research on data integration systems, exploring various mapping systems and languages, and their complex results and limitations. However no single technology suffices for all needs of an enterprise and these needs keep changing with growing data volumes and changing business needs. Consequently, enterprises end up tweaking their integration systems continua...

AI Technical Summary

Benefits of technology

The present invention provides a system and method for easily migrating data across different integration platforms. This is done by using a set of conceptual modeling tools and a set of primitive components that conform to a reference model driven architecture. The method involves modeling data at multiple levels of abstraction, using representational formalisms to capture relationships and rules, specifying the set of components needed for purpose-specific data integrations, mapping between source and global schemas, translating queries, and executing data flow graphs. The technical effects of this invention include improved efficiency and ease of migration across different integration platforms.

Problems solved by technology

However no single technology suffices for all needs of an enterprise and these needs keep changing with growing data volumes and changing business needs.

This consumes a lot of effort—by some estimates as much as 40% of all IT efforts in an enterprise.

One reason why this consumes so much effort is the rigidity of the available integration technologies.

Once a solution is implemented in one of these technologies, moving to another is like implementing the entire solution afresh which requires large amount of time, efforts and computational resources.

As a result, people end up building ad-hoc, quick-fix solutions, which over time leads to data fragmentation and inconsistencies.

Keeping these fragments synchronized to avoid inconsistencies puts a lot of strain on these systems.

The principal reason for lack of interoperability among tools of the same category (say ETL) is the lack of common reference architecture across tool implementations.

The principal reason for lack of interoperability across categories (say ETL and EII) is that their specifications are too close to the implementation platform i.e. they are not at a level of abstraction that allows their semantics to be mapped easily.

Moreover, the data inter-operability problem arises from the fact that data, even within a single domain of application, is available at many different sites, in many different schemas, and even in different data models.

Images