Denoising Implicit Feedback for Cold-start Recommendation

2026-06-17

AI Summary

This work addresses the challenge of noisy implicit feedback—such as clickbait and position bias—in cold-start recommendation scenarios. To tackle this issue, the authors propose DIF, a model-agnostic denoising framework that specifically targets cold-start items. DIF leverages content-similar warm items to generate high-confidence pseudo-labels and dynamically refines these labels at the sample level by integrating relative entropy with an adaptive correction mechanism conditioned on the item’s cold-start status. Extensive experiments on real-world datasets demonstrate that DIF significantly enhances cold-start recommendation performance. The method has been deployed on Kuaishou, a short-video platform serving hundreds of millions of users, yielding substantial improvements across multiple key business metrics.

Abstract

Implicit feedback is widely used in recommender systems due to its accessibility and generality, yet it usually presents noisy samples (e.g., clickbait, position bias). Meanwhile, recommenders inevitably face the item cold-start problem due to the continuous influx of new items. We identify that cold items are more prone to noisy samples due to the aforementioned factors, and researchers often overlook the significance of denoising implicit feedback for cold items. Previous denoising studies usually identify noisy samples based on heuristic patterns, such as higher loss values, and mitigate noise through sample selection or re-weighting. However, these methods have limited adaptability and are ineffective in cold-start scenarios. To achieve denoising implicit feedback for cold-start recommendation, we propose a model-agnostic denoising method called DIF. First, user preferences for content remain stable, which allows us to infer pseudo-labels indicating whether a user is interested in a cold item through content-similar warm items. Furthermore, to improve pseudo-label accuracy, we model the confidence of pseudo-labels based on the content similarity between the cold item and warm items, and then aggregate multiple pseudo-labels for each sample. Finally, we explicitly estimate the uncertainty of the noisy sample label by considering its relative entropy and the cold-start status of the item, which adaptively guides the role of pseudo-labels to correct the noisy labels at the sample level. DIF's superiority is supported by both theoretical justification and extensive experiments on real-world datasets. The method has been deployed on a billion-user scale short video application Kuaishou and has significantly improved various commercial metrics within cold-start scenarios.

Problem

• implicit feedback
• denoising
• cold-start recommendation
• noisy samples
• item cold-start

Innovation

• denoising implicit feedback
• cold-start recommendation
• pseudo-labeling
• content similarity
• uncertainty estimation