At one level, Analytics is just another Data Consumer and should follow all the same DM rules and requirements as any other Data Consumer in your organization. There are, however, some very special factors to consider when managing data for advanced analytics. In this overview, we’ll talk about:

• How the need for aggregation, anonymization, and obfuscation imposes special data quality rules.
• How Technology and DM need extra degrees of integration for analytics. Are Model Explainability and Model Management the purview of technology or DM…or both? How do we define version in the context of machine learning? Is it a form of software management or data management? 
• How is bias managed? 
• How about ethics? 
• What other special questions do we need to ask about our data management practices before we send our data to the analytics team?

In short, when discussing analytics, the very boundaries of what Data Management means can become a bit fuzzy. When we consider the various factors that go into defining these boundaries, it is less important where we draw those boundaries than that we are clear that all the factors have been accounted for.  

Speaker: David Kowalski

David Kowalski, Ph.D., has spent his entire professional life showing organizations how best to manage and govern their data. David is currently Principal Consultant at Ortecha, a data management executive advisory firm with offices in London, Nashville, and Princeton. He has worked with Data Management executives at large and mid-sized corporations around the world to assess the efficacy of their Data Management practices and help to devise strategies & policies to improve the efficiency, effectiveness, and reliability of these initiatives while mitigating risk exposure around these practices. An industry thought leader, David appears regularly at conferences across the country and is a very active contributing member of the Enterprise Data Management Council.

You can connect with David on LinkedIn or email him at Ortecha.