Nancy D. Griffeth
Nancy Davis Griffeth is an American computer scientist notable for approaches to the feature interaction problem. In 2014, she is a professor at Lehman College of The City University of New York and is modelling biological systems in computational biology.
Early life
Griffeth was born in Oak Park, Illinois and lived in Laurel, Mississippi and Memphis, Tennessee as a child. She received a bachelor's degree from Harvard University, a master's degree from Michigan State University, and a PhD degree from the University of Chicago.Career
Griffeth did seminal work in the feature interaction problem as a founding organizer of the feature interaction workshops and co-author of one of the most cited papers in feature interactions, "A Feature Interaction Benchmark for IN and Beyond."The feature interaction problem is a software problem that arises when one feature interacts with another in such a way that it changes what the feature does. This can cause serious issues for developers and users of the software. The problem was first documented as features were added to telecommunications systems. If new features on a telecommunications network were either undetected or unwanted, they could cause confusion and dissatisfaction among customers if not handled properly.
Griffeth also researched the related problem of how to test networks to see how well they work together, called "interoperability".
She worked at the Next Generation Networking Lab at Lucent Technologies where she designed and built tools to test interoperability of Voice-over-IP networks, which included conformance testing for MeGaCo media gateways and controllers. For this purpose, she also researched how to model protocols.
In addition, she studied virtual node layers regarding Mobile Ad Hoc Networks or MANETs. In the 1990s, she patented methods to protect databases against hackers trying to deduce confidential attributes. Her research has included distributed databases, simulations, concurrency and recovery controls, database design issues, performance modeling, and other issues.