AdaF^2M^2: Comprehensive Learning and Responsive Leveraging Features in Recommendation System

2025-01-27

AI Summary

To address the issue in recommender systems where long-tailed item distributions cause models to over-rely on ID features while neglecting intrinsic user/item attributes, this paper proposes AdaF²M²—a model-agnostic, adaptive feature modeling framework. Our method introduces two key innovations: (1) a novel feature-masking-driven multi-forward augmentation training mechanism that enforces robust learning of non-ID meta-features; and (2) a state-responsive lightweight adapter enabling dynamic, synergistic modeling of ID and meta-features. These components collectively enhance model generalization and noise resilience. Extensive A/B tests across multiple scenarios at Douyin Group demonstrate statistically significant improvements: +1.37% in user active days and +1.89% in app session duration. AdaF²M² has been successfully deployed at scale in both retrieval and ranking stages of production recommendation pipelines.

Technology Category

• Data Mining & Knowledge Management: Recommender Systems
• Machine Learning: Feature Construction/Reformulation
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Feature modeling, which involves feature representation learning and leveraging, plays an essential role in industrial recommendation systems. However, the data distribution in real-world applications usually follows a highly skewed long-tail pattern due to the popularity bias, which easily leads to over-reliance on ID-based features, such as user/item IDs and ID sequences of interactions. Such over-reliance makes it hard for models to learn features comprehensively, especially for those non-ID meta features, e.g., user/item characteristics. Further, it limits the feature leveraging ability in models, getting less generalized and more susceptible to data noise. Previous studies on feature modeling focus on feature extraction and interaction, hardly noticing the problems brought about by the long-tail data distribution. To achieve better feature representation learning and leveraging on real-world data, we propose a model-agnostic framework AdaF^2M^2, short for Adaptive Feature Modeling with Feature Mask. The feature-mask mechanism helps comprehensive feature learning via multi-forward training with augmented samples, while the adapter applies adaptive weights on features responsive to different user/item states. By arming base models with AdaF^2M^2, we conduct online A/B tests on multiple recommendation scenarios, obtaining +1.37% and +1.89% cumulative improvements on user active days and app duration respectively. Besides, the extended offline experiments on different models show improvements as well. AdaF$^2$M$^2$ has been widely deployed on both retrieval and ranking tasks in multiple applications of Douyin Group, indicating its superior effectiveness and universality.

Problem

• Recommendation Systems
• Long-tail Distribution
• Model Stability

Innovation

• AdaF^2M^2
• Feature Modeling
• Long-tail Distribution Adaptation