Stephen Bonner

Causal embeddings for recommendation

2018

Many current applications use recommendations in order to modify the natural user behavior, such as to increase the number of sales or the time spent on a website. This results in a gap between the …