Generally, when I read something about service-oriented architecture and integration, it's focused on how SOA can address integration. There's some debate over how much your SOA efforts should focus on integration -- it's an easy selling point, and possibly a good way to achieve an ROI, but many SOA experts say you can short-change yourself by focusing too much on integration.
That's old hat, right? But this month, I ran across a different kind of white paper about SOA and integration. The piece was published by BP Trends and written by Dr. Gopala Krishna Behara, a senior enterprise architect for Wipro Consulting Services.
I'll be honest with you -- I debated whether to refer to this white paper. Frankly, I found it a bit hard to follow and sometimes I felt the author lost focus from his main point. But I decided to point it out to you for a few reasons:
One of the topics he covers briefly is semantic integration, which is something I've been researching lately. Semantic integration isn't a huge topic right now -- in fact, it's still very much the domain of large companies. But I suspect you'll be hearing more about it in the coming years as vendors acquire semantic companies and start to incorporate those functions into their products.
Recently I interviewed Vickie Farrell, Neoview Product Management, of Hewlett-Packard's Business Information Optimization, about semantic integration. Farrell explained how semantic technologies can help with integration and offered some real-world applications, adding that most CIOs will encounter semantic integration technologies through other products:
We are starting to see more and more large mainstream companies acquire the technology and bake it into their products. That's probably how most people will see it, is baked it into those products.
Behara's paper managed to help me fill in a few gaps about metadata and integration, which in turn helped me understand semantic integration a bit better.
Of course, providing a history of integration was not Behara's goal in the "SOA and Integration" white paper. According to the conclusion, he was attempting to show how integration technologies can be leveraged for SOA.
As I said, I found that thesis a bit hard to follow, but enterprise architects probably will find it much clearer. For the more casual integration reader, however, Behara has managed to write a useful resource that might address those nagging integration questions you couldn't find answered anywhere else.