At the recent Data Governance and Information Quality Conference in San Diego I came across a new breed of innovators. These innovators are not trying to solve new problems but are finding innovative ways to solve existing problems. While the details of these problems vary across organizations these innovators all had one common solution theme or recommended practice –start with a small data governance initiative, prove its value and then expand the program.
The term ‘governance’ itself sometimes carries a negative connotation but the fact is that every organization does it to some degree. It may not be called governance or there may be no name for it at all. However every day someone in an organization shares information with colleagues, across departments, and software systems; correspondingly there are people/systems on the consuming end. It is an established fact that data is an enterprise asset and from what I heard at the DGIQ conference the ad hoc management of this asset leads to the following types of complications:

  • An executive realizing “I am only 40% sure of the report I am signing.
  • The CEO gets two reports, both backed by data with conflicting statements.
  • There is duplicate information all over the enterprise, for example, different systems each using their own taxonomies.
  • It takes months to gather data for a particular report that business is demanding. The ball is tossed between business and IT. Is it too late for the business when the report finally becomes available?

As human beings we think in terms objects and connections. Consider the human learning process. We have some knowledge in our brain and when we come across new information we are able to augment the knowledge we already have with this additional information about the object involved. How are we able to do that? You either know a fact about an object or you don’t. When you come across that fact, you add it to your knowledge base for that object.

Why doesn’t the information gathered in enterprises behave that way? The fact is that it can. One way to do this is in using graphs to collect and connect information. Imagine that all the information in your enterprise were stored as a connected graph with edges and nodes. Revisit the challenges above with the graph representation in your mind:

  • Reports link back to the terms they refer to and the data points which were used to gather the report.
  • Data analysts gathering data for the reports have a complete view of specific data elements at hand. They are linked to taxonomies and business glossaries and you can view a term’s current usage, you can see what software systems use it and how. You can see all the metadata attached to the term.
  • When needed, taxonomies can be linked together with the easy addition of simple statements such as: Term A (in taxonomy 1) is the same as term B (in taxonomy 2).
  • Data assets are linked to the business terms and can be assigned to people in the organization for responsibility and monitoring (e.g. RACI roles). Business users themselves can now get a high level picture of the data.

When these connections exist at the data layer there are some neat things you could do, such as:

  • Query across your entire dataset ecosystem.
  • If the UI is model driven, the changes in the data are reflected in the UI with no or minimal changes.
  • The data is essentially schema less and is easily extensible.

To address the needs of the data governance innovators in numerous organizations, who as noted want to use an incremental, evolvable approach to support, expand and integrate their data governance solutions, TopQuadrant created a new type of agile data governance solution for today’s dynamic enterprises – TopBraid Enterprise Data Governance™ . (TopBraid EDG™). Using standards-based graph technologies, TopBraid EDG supports integrated data governance across the ever growing numbers and types of data assets and governance needs – because connections are important.

<< Click on image below for larger view >> It provides flexibility and extensibility for data governance —enabling stakeholders to start with a simple, focused area of governance, e.g. managing a business glossary, and incrementally expanding to comprehensive governance of critical business assets including reference data, crosswalks, ontologies, taxonomies, content tag sets, data, policy and lineage models and other business, technical and operational metadata.

TopBraid EDG combines years of experience in the enterprise information management space and augments the enterprise-tested platform features of TopBraid Enterprise Vocabulary Net with data governance business models and use cases.