Operational Considerations in NLP

Proponents of exploratory data analysis assert that allowing the data to group naturally without any imposed framework is the most robust and unbiased approach. However, in a business setting there may be times when NLP yields results that are unactionable or confusing due to the difference between the traditional semantic meaning and the business use case. We examine the same data under two different frameworks and provide practical guidance on when to let the data do the talking, and when to let the business speak for itself.
Who is this presentation for?
Data Science Practitioners as well as Operational Professionals (Those who interpret call data)
Prerequisite knowledge:
None needed.
What you'll learn?
How to identify whether to let the data do the talking or to provide a business structure for analysis.

Amy Neftzger has worked in healthcare analytics for over two decades. As an applied researcher, her focus is on utilizing data science to improve operational processes and healthcare outcomes. Her experience includes working for vendors, Medicare Approved Contractors, and health plans, as well as serving on industry committees and special workgroups for The Disease Management Association of America, URAC, The Gallup Well-Being Index, and DaVinci FHIR (HL7). She currently works for United Healthcare as Senior Director of Data Science, UHC Provider Intelligence managing a team supporting data science and insights for provider network operations.