# Learning Robust Recommender from Noisy Implicit Feedback

arXiv: https://arxiv.org/abs/2112.01160 · HTML: https://mlsanity.com/paper/2112.01160/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 61 references that are in the map.
- First posted on arXiv 2 Dec 2021.

## Builds on (most cited first, 30 of 31)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Neural Collaborative Filtering](https://mlsanity.com/paper/1708.05031/) (2017)
- [Co-teaching: Robust Training of Deep Neural Networks with Extremely Noisy Labels](https://mlsanity.com/paper/1804.06872/) (2018)
- [A Closer Look at Memorization in Deep Networks](https://mlsanity.com/paper/1706.05394/) (2017)
- [Neural Graph Collaborative Filtering](https://mlsanity.com/paper/1905.08108/) (2019)
- [MentorNet: Learning Data-Driven Curriculum for Very Deep Neural Networks on Corrupted Labels](https://mlsanity.com/paper/1712.05055/) (2017)
- [Ups and Downs: Modeling the Visual Evolution of Fashion Trends with One-Class Collaborative Filtering](https://mlsanity.com/paper/1602.01585/) (2016)
- [How does Disagreement Help Generalization against Label Corruption?](https://mlsanity.com/paper/1901.04215/) (2019)
- [Unbiased Scene Graph Generation from Biased Training](https://mlsanity.com/paper/2002.11949/) (2020)
- [Variational Autoencoders for Collaborative Filtering](https://mlsanity.com/paper/1802.05814/) (2018)
- [Counterfactual VQA: A Cause-Effect Look at Language Bias](https://mlsanity.com/paper/2006.04315/) (2020)
- [Distilling Causal Effect of Data in Class-Incremental Learning](https://mlsanity.com/paper/2103.01737/) (2021)
- [Managing Popularity Bias in Recommender Systems with Personalized Re-ranking](https://mlsanity.com/paper/1901.07555/) (2019)
- [Denoising Implicit Feedback for Recommendation](https://mlsanity.com/paper/2006.04153/) (2020)
- [Causal Embeddings for Recommendation](https://mlsanity.com/paper/1706.07639/) (2017)
- [Unbiased Recommender Learning from Missing-Not-At-Random Implicit Feedback](https://mlsanity.com/paper/1909.03601/) (2019)
- [IRGAN: A Minimax Game for Unifying Generative and Discriminative Information Retrieval Models](https://mlsanity.com/paper/1705.10513/) (2017)
- [Modeling User Exposure in Recommendation](https://mlsanity.com/paper/1510.07025/) (2015)
- [Controlling Fairness and Bias in Dynamic Learning-to-Rank](https://mlsanity.com/paper/2005.14713/) (2020)
- [Unbiased Learning to Rank with Unbiased Propensity Estimation](https://mlsanity.com/paper/1804.05938/) (2018)
- [GCN-Based User Representation Learning for Unifying Robust Recommendation and Fraudster Detection](https://mlsanity.com/paper/2005.10150/) (2020)
- [Deconfounded Recommendation for Alleviating Bias Amplification](https://mlsanity.com/paper/2105.10648/) (2021)
- [Clicks can be Cheating: Counterfactual Recommendation for Mitigating Clickbait Issue](https://mlsanity.com/paper/2009.09945/) (2020)
- [Should Graph Convolution Trust Neighbors? A Simple Causal Inference Method](https://mlsanity.com/paper/2010.11797/) (2020)
- [Review of The Book of Why: The New Science of Cause and Effect](https://mlsanity.com/paper/2003.11635/) (2020)
- [Sampler Design for Implicit Feedback Data by Noisy-label Robust Learning](https://mlsanity.com/paper/2007.07204/) (2020)
- [Learning to Recommend with Multiple Cascading Behaviors](https://mlsanity.com/paper/1809.08161/) (2018)
- [LambdaOpt: Learn to Regularize Recommender Models in Finer Levels](https://mlsanity.com/paper/1905.11596/) (2019)
- [To Model or to Intervene: A Comparison of Counterfactual and Online Learning to Rank from User Interactions](https://mlsanity.com/paper/1907.06412/) (2019)
- [Empowering Language Understanding with Counterfactual Reasoning](https://mlsanity.com/paper/2106.03046/) (2021)

## Built on by (most cited first, 1 of 1)

- [CAdam: Confidence-Based Optimization for Online Learning](https://mlsanity.com/paper/2411.19647/) (2024)

## Abstract (arXiv)

The ubiquity of implicit feedback makes it indispensable for building recommender systems. However, it does not actually reflect the actual satisfaction of users. For example, in E-commerce, a large portion of clicks do not translate to purchases, and many purchases end up with negative reviews. As such, it is of importance to account for the inevitable noises in implicit feedback. However, little work on recommendation has taken the noisy nature of implicit feedback into consideration. In this work, we explore the central theme of denoising implicit feedback for recommender learning, including training and inference. By observing the process of normal recommender training, we find that noisy feedback typically has large loss values in the early stages. Inspired by this observation, we propose a new training strategy named Adaptive Denoising Training (ADT), which adaptively prunes the noisy interactions by two paradigms (i.e., Truncated Loss and Reweighted Loss). Furthermore, we consider extra feedback (e.g., rating) as auxiliary signal and propose three strategies to incorporate extra feedback into ADT: finetuning, warm-up training, and colliding inference. We instantiate the two paradigms on the widely used binary cross-entropy loss and test them on three representative recommender models. Extensive experiments on three benchmarks demonstrate that ADT significantly improves the quality of recommendation over normal training without using extra feedback. Besides, the proposed three strategies for using extra feedback largely enhance the denoising ability of ADT.