Thursday, September 24, 2009

How Semantic Technology helps Data Strategy? The discipline of Business Semantics!

Data Strategy has again become a "hot topic" in the enterprise. The two things which are of utmost importance in this area or require painstaking work are:

1) Data Standardizaton - Agreeing what to call things, which things are equivalent
2) Data Governance - Who "owns" the data, who is the steward, who gets to change things, who is responsible for quality

If you really think about it then more time is spent on analyzing the quality of data and doing reconciliation then anything else. Bottomline is that there has to be a common language between business and IT. Business semantics or developing  a dictionary based solution is the key here. You can call this business dictionary as a knowledge base or a an ontology based solution.

In my view, an “Enterprise Ontology” in the Semantic world is equivalent to an Enterprise Data Model in the Relational world. Developing an ontology is similar to creating a data model. However, one important difference is that ontologies allow you to seamlessly move from logical design to physical implementation. Also if you have established a business dictionary using the Web Ontology Language (OWL), you have encoded it in machine-processable as well as human-readable form.

Relational databases (and the applications that use them) behave as if every row in a table is unique. So when a customer database contains "Microsoft” and “Microsoft Inc.,” the system behaves as if they are two different customers. In the world of ontologies, we should be able to simply declare the equivalence of the two instances and be done with it.

Business Semantic is a fresh market and there are very few case studies out there. This market will only grow in the long run as using semantic technology approach for ensuring data lineage and data quality is the best possible approach from flexibility and maintability perspective. There is a start up in Belgium called Collibra which is actively working on it using purely semantic approach.  

Reblog this post [with Zemanta]

No comments:

Post a Comment