When you look at what you need to be successful in SOA infrastructure, data is really fundamental. While it's important, of course, to have governance and visibility into all the systems that you build in a distributed SOA world, it's also important to have SOA governance and management over the information (or data) that's flowing through the systems. Semantic integration and semantic interoperability ensure that you have that kind of governance by creating a common information model that ensures that all the information passed between all your various systems, and services, within your SOA is consistent. With Semantic Integrator (SI), you can create rules and consistency across all those different applications, services and models.
In fact, I would argue (and I did mention this in my last post entitled, SOA - What's there to be afraid of? ) that one of the biggest reasons why SOA implementations fail is because of too little concern over the actual data issues. Let’s face it, connection is one thing but actual communication and data exchange is a completely different matter. And for your service-oriented architecture (SOA) to be effective, you have to get all the way to actually communicating the data/information that is required to make a transaction or event successful.
SOA what? Semantic data integration is central to any effective SOA infrastructure implementation.