Dual Unbiased Recommender Learning for Implicit Feedback
Authors:
Open Access
- 11 July 2021
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1647-1651
Abstract
Unbiased recommender learning has been actively studied to alleviate the inherent bias of implicit datasets under the missing-not-at-random assumption. Existing studies solely address the bias of positive feedback but do not account for the bias of missing feedback, which heavily affects their sub-optimal performance gains. This paper proposes a dual recommender learning framework that simultaneously eliminates the bias of clicked and unclicked data. Specifically, the proposed loss function adopts two propensity weighting to effectively estimate the true positive and negative preferences from clicked and unclicked data. We also prove that the proposed loss function converges to the ideal loss function for both clicked and unclicked data. Because of the model-agnostic property, it can be applied to any existing unbiased learning models. Experimental results show that the proposed method outperforms state-of-the-art unbiased models up to 5.54-24.56% for [email protected] on three datasets.Keywords
Funding Information
- National Research Foundation ofKorea (NRF) (No. NRF-2018R1A5A1060031)
- Institute of Informa-tion & communications Technology Planning & evaluation (IITP) (No. 2019-0-00421)
This publication has 11 references indexed in Scilit:
- Unbiased Implicit Recommendation and Propensity Estimation via Combinational Joint Learning2020
- Unbiased Pairwise Learning from Biased Implicit Feedback2020
- Attribute-based Propensity for Unbiased Learning in Recommender Systems2020
- Asymmetric Tri-training for Debiasing Missing-Not-At-Random Explicit Feedback2020
- Unbiased Recommender Learning from Missing-Not-At-Random Implicit Feedback2020
- Unbiased offline recommender evaluation for missing-not-at-random implicit feedback2018
- Should I Follow the Crowd?2018
- Learning to Rank with Selection Bias in Personal Search2016
- Item popularity and recommendation accuracy2011
- Training and testing of recommender systems on data missing not at random2010