7 Steps to Smarter Integration
Sometimes, change can be worthwhile. The key is knowing what's worth pursuing and what's not.
Maybe it's because I'm a writer, but the Semantic Web and Linked Data in particular are fascinating to me. From the URL-based naming for data to the way you can find Linked Datasets online, it's bound to change integration for IT. So, I try to follow it.
For whatever reason, I only find one or two pieces a year that really speak to Linked Data's application in the enterprise. Maybe I'm looking in the wrong places, I don't know.
But this month, I found what is, to date, the Best Piece Thus Far for addressing the question: What can Linked Data do for enterprise IT?
The Semantic Web Journal published last week an excellent research paper, "Linked Data in Enterprise Information Integration." The 17-page document explains how Linked Data can be used for typical integration challenges. Oh, and you can download a PDF of it for free.
I think it's a must-read for visionary CIOs and high-level IT managers, as well as anyone who deals with data management, the information cycle, or data modeling.
Business users may find it challenging, depending upon your comfort with IT acronyms and knowledge of technology. Even so, you can get the general idea in the first few pages.
The paper points out that data integration solutions in the past decade have focused on XML, Web services and Service Oriented Architectures. While that's worked well for transaction processing, it hasn't actually resolved the challenges of data integration in the enterprise, the paper argues. The writers then go on to explain how the "Linked Data paradigm" is a better option for dealing with enterprise data integration.
This isn't theoretical: It's very specific in the discussion of challenges and even how you can start building what the writers call the "enterprise data web."
There are even real examples. Microsoft and Google are among those companies starting to move into using the "Linked Data paradigm for the integration of their thousands of distributed information systems." What's more, the piece adds, these companies will "aim to establish Enterprise Knowledge Bases (similar to what Freebase now is for Google) as hubs and crystallization points for the vast amounts of structured data and knowledge distributed in their data intranets."
The starting point: Your existing enterprise taxonomies, which will serve as the "nucleus of linking and integration hubs." It makes sense, because these taxonomies already reflect many of the important terminology and culture of the organization.
But the paper explains you will need to add datasets from the Linked Open Data Web to make these comprehensive enterprise knowledge bases.
The payoff? You can use your enterprise knowledge base for:
It sounds almost like a some sort of virtual intergalactic Starbase for data, with all the data ships coming in to be updated and integrated with all the other data hubs. It's definitely exciting - although let's hope my rate of finding pertinent articles on the topic isn't any indication of the timeframe for real adoption. Otherwise, we may be trying Linked Data integration while on an actual Starbase.