Five E's: Reflecting on the Design of Recommendations DUPE_TO_DELETE

Keynote at ACM RecSys 2018

RecSys 2018 opening keynote reflecting on recommendation design through five E's: Explainable, Equitable, Ethical, Expedient and Exigent. Uses psychology of information processing and cautionary cases (filter bubbles, discriminatory recommendations, poor data quality) to ask what researchers, developers and designers can do concretely.

People

Elizabeth Churchill