Katherine Livins: Feature Selection For Human Recommenders

Agenda:

• 18.30: Doors Open

• 19.00: Main Presentation

• 19:45: Q/A

• 20.00: Mingling

Abstract

The increasing popularity of human-in-the-loop recommendation algorithms brings a question about divisions of labor. To the point, these algorithms often rely on computers to perform repetitive and large-scale calculations, and humans to consider unstructured data and to make holistic decisions. As a result it is rare that the two systems will process an identical feature set. Feature selection methods with computers are well established, however feature selection with humans is relatively unexplored. This talk will outline how experimental methods (borrowed from the behavioral sciences) can be used to this end, along with how the human component of the system may be improved as a result.

Bio

Katherine is a Data Scientist at Stitch Fix, working on the Human Computation team under Algorithms. Stitch Fix uses a hybrid recommendation system, which involves both computer algorithms and human ‘stylists’. Katherine works on understanding stylist behavior through data and experimentation, and then leverages what she learns to optimize their workflow. Before Stitch Fix, Katherine completed a PhD in Cognitive and Information Science where she researched how to shape human behavior and reasoning through the manipulation of attentional cues.