Dual Unbiased Recommender Learning for Implicit Feedback

Lee, Jae-woong; Park, Seongmin; Lee, Jongwuk

doi:10.1145/3404835.3463118

Cited by 13 publications

(14 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Zhang et al [25] proposes to co-learn CTR, CVR and conducted debiasing within a multi-task learning framework. Recently, Lee et al [11] proposes a dual learning framework that simultaneously eliminates the confounding effect in clicked and unclicked data.…”

Section: Related Workmentioning

confidence: 99%

ESCM$^2$: Entire Space Counterfactual Multi-Task Model for Post-Click Conversion Rate Estimation

Wang,

Chang,

Liu

et al. 2022

Preprint

View full text Add to dashboard Cite

Accurate estimation of post-click conversion rate is critical for building recommender systems, which has long been confronted with sample selection bias and data sparsity issues. Methods in the Entire Space Multi-task Model (ESMM) family leverage the sequential pattern of user actions, i.e., 𝑖𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 → 𝑐𝑙𝑖𝑐𝑘 → 𝑐𝑜𝑛𝑣𝑒𝑟𝑠𝑖𝑜𝑛 to address data sparsity issue. However, they still fail to ensure the unbiasedness of CVR estimates. In this paper, we theoretically demonstrate that ESMM suffers from the following two problems:(1) Inherent Estimation Bias (IEB), where the estimated CVR of ESMM is inherently higher than the ground truth; (2) Potential Independence Priority (PIP) for CTCVR estimation, where there is a risk that the ESMM overlooks the causality from click to conversion. To this end, we devise a principled approach named Entire Space Counterfactual Multi-task Modelling (ESCM 2 ), which employs a counterfactual risk miminizer as a regularizer in ESMM to address both IEB and PIP issues simultaneously. Extensive experiments on offline datasets and online environments demonstrate that our proposed ESCM 2 can largely mitigate the inherent IEB and PIP issues and achieve better performance than baseline models.

show abstract

Section: Related Workmentioning

confidence: 99%

ESCM$^2$: Entire Space Counterfactual Multi-Task Model for Post-Click Conversion Rate Estimation

Wang,

Chang,

Liu

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Later, UBPR [29] is proposed to extend the pointwise model in Rel-MF to a pairwise version. UEBPR [12] and DU [21] introduce neighborhood-based explainability and unclicked data reweighting to the plain IPS methods, respectively. AutoDebias [6] combines IPS with data imputation and adopts a meta-learning algorithm to learn the optimal debiasing configurations on a small uniform data.…”

Section: Related Workmentioning

confidence: 99%

“…How to mitigate the bias and perform unbiased estimation of user preference has become one central theme in recommendation [7]. One prevalent solution is Inverse Propensity Scoring (IPS) [12,21,29,30], which reweights each data sample by the inverse of its propensity score (i.e., the exposure probability). Despite that IPS is unbiased in theory, it suffers from practical limitations: (1) it is challenging to accurately estimate the propensity score of each sample, since the exposure mechanism is seldom known [31]; and (2) the reweighted loss usually exhibits a high variance especially for implicit feedback [29,30,32], which implies that the losses of individual samples fluctuate heavily from their expected values.…”

Section: Introductionmentioning

confidence: 99%

Cross Pairwise Ranking for Unbiased Item Recommendation

Wan

Wang

et al. 2022

Proceedings of the ACM Web Conference 2022

View full text Add to dashboard Cite

Most recommender systems optimize the model on observed interaction data, which is affected by the previous exposure mechanism and exhibits many biases like popularity bias. The loss functions, such as the mostly used pointwise Binary Cross-Entropy and pairwise Bayesian Personalized Ranking, are not designed to consider the biases in observed data. As a result, the model optimized on the loss would inherit the data biases, or even worse, amplify the biases. For example, a few popular items take up more and more exposure opportunities, severely hurting the recommendation quality on niche items -known as the notorious Mathew effect.In this work, we develop a new learning paradigm named Cross Pairwise Ranking (CPR) that achieves unbiased recommendation without knowing the exposure mechanism. Distinct from inverse propensity scoring (IPS), we change the loss term of a sample -we innovatively sample multiple observed interactions once and form the loss as the combination of their predictions. We prove in theory that this way offsets the influence of user/item propensity on the learning, removing the influence of data biases caused by the exposure mechanism. Advantageous to IPS, our proposed CPR ensures unbiased learning for each training instance without the need of setting the propensity scores. Experimental results demonstrate the superiority of CPR over state-of-the-art debiasing solutions in both model generalization and training efficiency. The codes are available at https://github.com/Qcactus/CPR. CCS CONCEPTS• Information systems → Retrieval models and ranking; Recommender systems.

show abstract

“…The critical issue is how to estimate the propensity score 𝜔 𝑢𝑖 from observed feedback (e.g., clicks). Previous studies [23,33,38,39,63] have developed several solutions to estimate 𝜔 𝑢𝑖 . First, [23,38,39] introduced a heuristic function to estimate item popularity without using an additional model.…”

Section: Introductionmentioning

confidence: 99%

“…Previous studies [23,33,38,39,63] have developed several solutions to estimate 𝜔 𝑢𝑖 . First, [23,38,39] introduced a heuristic function to estimate item popularity without using an additional model. Although intuitive, it focuses only on addressing item popularity and is thus, incapable of handling exposure bias caused by recommender models.…”

Section: Introductionmentioning

confidence: 99%

Bilateral Self-unbiased Learning from Biased Implicit Feedback

Lee¹,

Park²,

Lee³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Implicit feedback has been widely used to build commercial recommender systems. Because observed feedback represents users' click logs, there is a semantic gap between true relevance and observed feedback. More importantly, observed feedback is usually biased towards popular items, thereby overestimating the actual relevance of popular items. Although existing studies have developed unbiased learning methods using inverse propensity weighting (IPW) or causal reasoning, they solely focus on eliminating the popularity bias of items. In this paper, we propose a novel unbiased recommender learning model, namely BIlateral SElf-unbiased Recommender (BISER), to eliminate the exposure bias of items caused by recommender models. Specifically, BISER consists of two key components: (i) self-inverse propensity weighting (SIPW) to gradually mitigate the bias of items without incurring high computational costs; and (ii) bilateral unbiased learning (BU) to bridge the gap between two complementary models in model predictions, i.e., user-and item-based autoencoders, alleviating the high variance of SIPW. Extensive experiments show that BISER consistently outperforms state-of-the-art unbiased recommender models over several datasets, including Coat, Yahoo! R3, MovieLens, and CiteULike.

show abstract

Dual Unbiased Recommender Learning for Implicit Feedback

Cited by 13 publications

References 13 publications

ESCM$^2$: Entire Space Counterfactual Multi-Task Model for Post-Click Conversion Rate Estimation

ESCM$^2$: Entire Space Counterfactual Multi-Task Model for Post-Click Conversion Rate Estimation

Cross Pairwise Ranking for Unbiased Item Recommendation

Bilateral Self-unbiased Learning from Biased Implicit Feedback

Contact Info

Product

Resources

About