Denoising Implicit Feedback for Recommendation

Abstract

The ubiquity of implicit feedback makes them the default choice to build online recommender systems. While the large volume of implicit feedback alleviates the data sparsity issue, the downside is that they are not as clean in reflecting 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 critical importance to account for the inevitable noises in implicit feedback for recommender training. 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 training. We find serious negative impacts of noisy implicit feedback,i.e., fitting the noisy data prevents the recommender from learning the actual user preference. Our target is to identify and prune noisy interactions, so as to improve the quality of recommender training. 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 namedAdaptive Denoising Training(ADT), which adaptively prunes noisy interactions during training. Specifically, we devise two paradigms for adaptive loss formulation: Truncated Loss that discards the large-loss samples with a dynamic threshold in each iteration; and reweighted Loss that adaptively lowers the weight of large-loss samples. We instantiate the two paradigms on the widely used binary cross-entropy loss and test the proposed ADT strategies on three representative recommenders. Extensive experiments on three benchmarks demonstrate that ADT significantly improves the quality of recommendation over normal training.

References in corpus (6)

• BPR: Bayesian Personalized Ranking from Implicit Feedback
• Understanding deep learning requires rethinking generalization
• A Closer Look at Memorization in Deep Networks
• How does Disagreement Help Generalization against Label Corruption?
• To Model or to Intervene: A Comparison of Counterfactual and Online Learning to Rank from User Interactions
• LambdaOpt: Learn to Regularize Recommender Models in Finer Levels

Cited by in corpus (43)

• Self-supervised Graph Learning for Recommendation
• Advances and Challenges in Conversational Recommender Systems: A Survey
• A Tale of Two Graphs: Freezing and Denoising Graph Structures for Multimodal Recommendation
• Representation Learning with Large Language Models for Recommendation
• Clicks can be Cheating: Counterfactual Recommendation for Mitigating Clickbait Issue
• Deconfounded Recommendation for Alleviating Bias Amplification
• Robust Preference-Guided Denoising for Graph based Social Recommendation
• Contrastive Self-supervised Sequential Recommendation with Robust Augmentation
• Denoising Self-attentive Sequential Recommendation
• Self-Guided Learning to Denoise for Robust Recommendation
• Market2Dish: Health-aware Food Recommendation
• Learning Robust Recommenders through Cross-Model Agreement
• Neighborhood-Enhanced Supervised Contrastive Learning for Collaborative Filtering
• AutoDenoise: Automatic Data Instance Denoising for Recommendations
• Knowledge-refined Denoising Network for Robust Recommendation
• Improving Implicit Feedback-Based Recommendation through Multi-Behavior Alignment
• Augmented Negative Sampling for Collaborative Filtering
• Uncovering User Interest from Biased and Noised Watch Time in Video Recommendation
• What is Next when Sequential Prediction Meets Implicitly Hard Interaction?
• LightGNN: Simple Graph Neural Network for Recommendation
• Multi-Modal Multi-Behavior Sequential Recommendation with Conditional Diffusion-Based Feature Denoising
• Rethinking Collaborative Metric Learning: Toward an Efficient Alternative without Negative Sampling
• Generating Usage-related Questions for Preference Elicitation in Conversational Recommender Systems
• Unleashing the Power of Large Language Model for Denoising Recommendation
• Dual-interest Factorization-heads Attention for Sequential Recommendation
• Integrating Item Relevance in Training Loss for Sequential Recommender Systems
• Balancing User Preferences by Social Networks: A Condition-Guided Social Recommendation Model for Mitigating Popularity Bias
• Unlocking the Hidden Treasures: Enhancing Recommendations with Unlabeled Data
• LabelCraft: Empowering Short Video Recommendations with Automated Label Crafting
• Why People Skip Music? On Predicting Music Skips using Deep Reinforcement Learning
• Adversarial Collaborative Filtering for Free
• Brain Topography Adaptive Network for Satisfaction Modeling in Interactive Information Access System
• Shapley Value-driven Data Pruning for Recommender Systems
• Adapting Triplet Importance of Implicit Feedback for Personalized Recommendation
• Lower-Left Partial AUC: An Effective and Efficient Optimization Metric for Recommendation
• Learning Robust Recommender from Noisy Implicit Feedback
• Recommendation Is a Dish Better Served Warm
• ConQueR: Contextualized Query Reduction using Search Logs
• Knowledge Graph-enhanced Sampling for Conversational Recommender System
• Empowering Denoising Sequential Recommendation with Large Language Model Embeddings
• Improved Personalized Headline Generation via Denoising Fake Interests from Implicit Feedback
• Enhancing Recommendation with Denoising Auxiliary Task
• Which Samples Should be Learned First: Easy or Hard?