[1706.07639] Causal Embeddings for Recommendation

Abstract: 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 final recommendation objective and the classical setup where recommendation candidates are evaluated by their coherence with past user behavior, by predicting either the missing entries in the user-item matrix, or the most likely next event. To bridge this gap, we optimize a recommendation policy for the task of increasing the desired outcome versus the organic user behavior. We show this is equivalent to learning to predict recommendation outcomes under a fully random recommendation policy. To this end, we propose a new domain adaptation algorithm that learns from logged data containing outcomes from a biased recommendation policy and predicts recommendation outcomes according to random exposure. We compare our method against state-of-the-art factorization methods, in addition to

3 mentions: @CriteoAILab@CriteoEng@shion_honda
Date: 2018/12/02 15:01

Related Entries

Read more FashionGen Challenge
2 users, 1 mentions 2019/03/22 00:48
Read more [1905.08527] CNNs found to jump around more skillfully than RNNs: Compositional generalization in se...
0 users, 1 mentions 2019/08/24 05:17
Read more [1806.07755] An empirical study on evaluation metrics of generative adversarial networks
0 users, 1 mentions 2019/08/27 15:47
Read more [1912.04958] Analyzing and Improving the Image Quality of StyleGANcontact arXivarXiv Twitter
0 users, 2 mentions 2020/02/12 20:21
Read more Meet Pandas: Group-wise Sampling | Hippocampus's Garden
0 users, 1 mentions 2020/10/13 02:21